How to Use Synthesia: A Beginner’s Guide (2026)

Introduction

Creating a professional video traditionally requires cameras, actors, microphones, editing software, and considerable production time. Synthesia simplifies this process by allowing users to create presenter-style videos with AI avatars and generated voices.

Users can begin with a written prompt, script, document, webpage, or uploaded file. Synthesia can then help organize the content into scenes, select layouts, add an AI presenter, generate narration, and create captions. Videos can also be translated into more than 160 languages, making the platform useful for training, education, product demonstrations, onboarding, marketing, and internal business communication.

Synthesia provides more than 240 stock avatars, along with options for creating personal avatars from a photo or recorded footage. Users can also create voiceover-only scenes when a visible presenter is unnecessary.

This beginner’s guide will explain how to:

• Create a Synthesia account and understand the dashboard
• Start a video from a prompt, template, script, webpage, or file
• Write and organize the video script
• Add, replace, and customize AI avatars
• Choose an AI voice and correct pronunciation
• Create a personal avatar and cloned voice
• Add scenes, layouts, backgrounds, images, video, and text
• Record your screen and add media
• Add captions, music, branding, and transitions
• Translate a video into another language
• Preview, generate, share, and download the finished video
• Manage projects, credits, and plan limits

Synthesia automates much of the production process, but users should still review every generated scene. Scripts, pronunciation, layouts, avatar placement, captions, and visual elements may require manual adjustments before the finished video is ready to publish.

Beginners should start with a short project so they can learn how the editor works without managing too many scenes. Once the basic workflow becomes familiar, the same tools can be used to create longer presentations, training courses, multilingual videos, and reusable branded content.

Create a Synthesia Account

Synthesia offers a free Basic plan that allows beginners to explore the editor, test AI avatars and voices, and create limited video content without entering a credit card. The current Basic plan includes 1,200 credits per month, which can be used for up to 10 minutes of generated video and 25 AI-generated visual assets.

Sign Up for Synthesia

1. Open the Synthesia website on a desktop computer.
2. Select Get Started.
3. Create an account using the available sign-up method.
4. Verify the email address when requested.
5. Enter the Synthesia application.
6. Complete any introductory setup questions.
7. Open the main workspace.

The free plan does not require payment information, so beginners can test the main creation process before choosing a paid subscription.

Choose the Right Plan

Beginners can start with the free Basic plan to learn how the editor, avatars, voices, scenes, and templates work. Synthesia also offers Starter, Creator, and Enterprise plans with different allowances and collaboration features.

The free plan is suitable for testing, but its restrictions may become noticeable when creating longer or more frequent videos. Paid plans add capabilities such as downloadable videos, removal of the Synthesia logo, more avatars, additional credits, and guest access.

Do not upgrade immediately unless the free plan prevents you from testing an important feature. Create a short practice video first, then compare the available plans based on the video length, avatar selection, export requirements, and monthly production volume you actually need.

Understand the Workspace

A workspace is the area where Synthesia videos are created, stored, and managed. It contains the projects and collaboration tools available under the account’s plan, license, and role.

Self-service plans are built around one primary user and may include guest access. Shared workspaces designed for broader team collaboration are currently associated with Synthesia’s Enterprise plan.

Beginners using Synthesia alone can usually remain in the default workspace. Separate workspaces become more useful when an organization needs to divide videos by team, department, client, or access level.

Review the Account Settings

To open the account settings:

1. Select your initials in the upper-right corner.
2. Choose Account Info.
3. Review your profile, email, password, notifications, and workspace settings.

Synthesia’s user settings allow users to update their name and profile image, change their password or email address, manage notifications, and review available integrations. API access is currently limited to Creator plans and above.

Confirm the Active Workspace

Users who belong to more than one workspace should confirm that the correct one is selected before creating a project. Videos, permissions, plan limits, and available assets may differ between workspaces.

The selected workspace appears near the upper-left area of the application. Choosing the correct workspace prevents a personal project from being created inside a company area or a team video from being stored in the wrong account environment.

Start With the Free Plan Carefully

Use the free account to learn how to create scenes, add an avatar, enter a script, select a voice, and generate a short test video. Avoid using the entire monthly allowance on the first experiment.

A short project of approximately 30 to 60 seconds is usually enough to test the basic workflow. Once the account is ready, the next step is learning how to navigate the Synthesia dashboard and locate the available video-creation tools.

Understand the Synthesia Dashboard

The Synthesia dashboard is the main area for creating videos, reopening projects, managing avatars and media, accessing templates, and checking workspace content. The exact navigation may vary depending on the account’s plan, role, and available Enterprise features.

Start a New Video

Select New Video or Create Video to begin a project. Synthesia may provide several starting options, including:

• Create with the AI Assistant
• Start from a blank video
• Use a template
• Import a PowerPoint, PDF, webpage, or other supported content
• Duplicate an existing video

The AI Assistant uses a chat-based workflow to help develop the script, choose a template, select a delivery style, and improve the video before or during editing. (help.synthesia.io)

Beginners who already have a finished script can use a blank project or template. The AI Assistant may be easier when the topic still needs to be organized.

Open My Videos

The My Videos area contains videos that belong to the individual user. These projects remain private unless they are shared or moved into a collaborative workspace location.

Select a video card to open it. The three-dot menu may provide actions such as duplicating, moving, renaming, sharing, or deleting the project.

Enterprise users may have both My Videos and a shared Workspace area. Videos stored in the workspace can be visible to other members, while projects under My Videos remain visible only to their owner. (help.synthesia.io)

Use the Workspace Area

A shared workspace stores projects intended for collaboration. Depending on the user’s permissions, members may be able to view, comment on, edit, move, or share its videos.

Confirm the active workspace before starting a project. Synthesia content is connected to the workspace where it was created, and a project stored in another workspace may not appear in the current one. (help.synthesia.io)

Users who belong to several workspaces can switch between them through the workspace selector.

Browse the Template Library

Open Library, then Templates, to browse reusable video designs. Templates provide prepared scenes, layouts, colors, text areas, and avatar positions that can be customized for a new project.

Synthesia includes its own templates, while eligible users and teams can also create reusable custom templates. A template can be created from scratch or saved from an existing video. (help.synthesia.io)

Choose a template based on the content rather than its appearance alone. A training template may provide better layouts for instructions, while a promotional template may emphasize products, statistics, and calls to action.

Open the Avatars Area

Select Avatars to browse available presenters and manage personal avatars. This area can also contain avatars that have been shared with the user or workspace.

Eligible users can select Create Avatar to begin creating a personal presenter from a photo or recorded video. Plan allowances and available creation methods vary. (help.synthesia.io)

You do not need to create a personal avatar before making a video. A stock avatar is sufficient for learning the editor.

Open the Assets Area

The Assets section may contain reusable resources such as:

• Brand Kits
• Uploaded media
• Voices
• Shared avatars
• Other workspace assets

Brand Kits allow eligible teams to store approved logos, colors, fonts, avatars, and visual themes. Synthesia currently places Brand Kits under Assets in the left-hand navigation. (help.synthesia.io)

Some asset-management and branding controls are limited to Enterprise accounts, so free or self-service users may not see every option.

Use the Media Library

The Media Library stores images and videos that can be reused across Synthesia projects. This is useful for company logos, product images, screenshots, demonstration footage, branded backgrounds, and frequently used visual assets. (help.synthesia.io)

Organizing reusable media in the library can prevent the same files from being uploaded separately for every project.

Search for Content

Use the search bar at the top of the dashboard to find videos, folders, and templates by title or description.

The search dropdown displays the strongest matches. Select See All Results to open a dedicated page with separate tabs and additional filters for videos, folders, and templates. (help.synthesia.io)

Use descriptive project titles so they remain easy to find later.

Create Folders

Folders help organize projects by client, department, course, platform, campaign, or publication status.

To create one:

1. Open My Videos or the appropriate workspace area.
2. Select Create Folder.
3. Enter the folder title.
4. Move projects into it by dragging them or using Move To.

Synthesia also supports subfolders and bulk project movement. (help.synthesia.io)

A simple beginner structure might include Drafts, Completed Videos, and Practice Projects.

Check Notifications and Account Settings

The account menu provides access to profile information, workspace settings, subscription details, and other account controls. Notifications may alert users about completed video generation, invitations, comments, shared content, or account activity.

Users should also monitor their remaining credits. Synthesia currently charges self-service accounts two credits for each generated second, meaning a one-minute video uses 120 credits. Regenerating an edited video may consume additional credits based on its updated duration. (help.synthesia.io)

Before creating a long project, check the available balance and use short previews or practice videos while learning.

Keep the Dashboard Organized

Rename new projects immediately, place completed work into folders, and confirm the correct workspace before editing. Avoid keeping several projects with names such as Untitled Video, because they become difficult to distinguish later.

Once the dashboard feels familiar, the next step is choosing the correct way to start a Synthesia video.

Choose How to Start a Synthesia Video

Synthesia provides several ways to begin a project. Users can generate a first draft with the AI Assistant, import a finished script, customize a template, build a video from a blank scene, or import an existing PowerPoint presentation.

The best starting method depends on how much content has already been prepared.

Start With the AI Assistant

Synthesia’s Assistant is a chat-based AI tool that can turn a prompt and supporting materials into a complete first draft. It can help write the script, organize scenes, choose visuals, select a template, and revise the project directly inside the editor.

To start with Assistant:

1. Select New Video.
2. Choose Start With Assistant.
3. Select Prompt to Video.
4. Describe the video you want to create.
5. Upload supporting materials when needed.
6. Choose the approximate duration.
7. Select a delivery style.
8. Choose a template.
9. Apply a Brand Kit when the option is available.
10. Select Create Video.

Assistant creates the first script, scenes, layouts, and visuals. The resulting project should still be reviewed inside the editor before it is generated.

Write a Detailed Prompt

A strong prompt should explain the topic, intended audience, purpose, and desired presentation style.

Instead of entering:

“Create an onboarding video.”

Use something more specific:

“Create a three-minute onboarding video for new remote customer-support representatives. Explain how to sign in, locate assigned tickets, respond to customers, and escalate technical problems. Use a professional but welcoming tone.”

Synthesia recommends including the topic, audience, and objectives because additional context generally produces a stronger first draft.

Include important requirements such as:

• The target length
• The intended platform
• Topics that must be covered
• The desired tone
• Any terminology the script should use
• Information that should not be added

Review all AI-generated claims before using them. Assistant may organize the material effectively while still producing inaccurate, generic, or unnecessary wording.

Attach Supporting Materials

Assistant can use supporting content alongside the prompt. Current supported sources include PDFs, PowerPoint presentations, Word documents, URLs, and multiple files uploaded together.

Supporting files can help Assistant follow existing company information instead of developing the complete video from a general prompt.

For example, users can provide:

• A product manual for a tutorial
• A company policy for an employee-training video
• A webpage for a product overview
• A presentation for a narrated lesson
• A written report for an executive summary

Explain how the material should be used. A file may contain more information than the final video needs, so identify the sections, facts, or recommendations that matter most.

Choose a Delivery Style

Assistant currently offers two primary delivery styles:

Cinematic creates shorter scenes with more frequent visual changes. It is better suited to dynamic explainers, promotional content, and videos intended to move quickly.

Presentation creates longer, denser scenes with fewer cuts. It may work better for training, detailed explanations, internal communication, and information-heavy material.

Choose based on the amount of information viewers need to process. A fast cinematic format may feel engaging, but it can make detailed instructions difficult to follow.

Import a Finished Script

Select Import Script when the wording is already complete and should not be rewritten by Assistant.

1. Open Start With Assistant.
2. Select Import Script.
3. Paste the script into the prompt field or attach a script file.
4. Choose the delivery style.
5. Select a template.
6. Apply an available Brand Kit when needed.
7. Click Create Video.

Synthesia uses the submitted script as written while creating scenes and visual layouts around it.

Review the script before importing it. Correcting the structure first is easier than reorganizing numerous generated scenes later.

Start With a Template

Templates provide prepared layouts, text areas, avatar positions, colors, and scene designs. They can help beginners produce a consistent video without designing every scene from the beginning.

Choose a template based on the video’s purpose. A training template may include layouts for steps, objectives, quotations, and summaries, while a promotional template may emphasize products, statistics, and calls to action.

Templates can be provided by Synthesia or created and published within an eligible workspace. Existing videos can also be converted into reusable templates.

After choosing a template:

1. Replace its sample text with the real script.
2. Change the avatar and voice.
3. Replace placeholder images and videos.
4. Remove scenes that are not needed.
5. Add new scenes when the content requires them.
6. Preview the complete video.

Do not force the script to fit every scene included in a template. Delete unnecessary layouts rather than adding filler content.

Start With a Blank Video

A blank project provides the most control. It is useful when the script is complete and the user wants to design each scene manually.

Begin with one empty scene, add the script, and then choose the avatar, background, text, media, and layout. Additional scenes can be added as the video develops.

Synthesia treats scenes similarly to slides in a presentation. Each scene can contain an avatar, text, images, and video clips organized around one main message.

A blank project requires more setup than Assistant or a template, but it avoids unnecessary AI-generated wording and layouts.

Import a PowerPoint Presentation

Synthesia can convert an existing PowerPoint presentation into a new video project.

1. Select New Video.
2. Choose Import PowerPoint.
3. Upload the presentation.
4. Review the available import settings.
5. Select Import.
6. Open the resulting project in the editor.

Text, shapes, images, and videos from the slides become editable Synthesia elements. Speaker notes are imported as the corresponding video script, and Synthesia attempts to detect the presentation’s fonts.

The direct PowerPoint importer currently supports .pptx files up to 1 GB or 150 slides. PowerPoint animations are not imported, so movement and transitions must be recreated inside Synthesia.

Review every imported slide for misplaced text, substituted fonts, unsupported formatting, or visual elements that no longer fit the video format.

Use the Legacy AI Video Assistant

Synthesia still provides its original template-based AI Video Assistant on supported paid plans. It can generate an outline, script, scenes, and layouts from prompts, scripts, files, or URLs. However, Synthesia states that this legacy tool will eventually be replaced by the newer chat-based Assistant.

The legacy assistant may still be useful for existing workflows, but new users should generally learn the current Assistant because it can continue refining the project directly inside the editor.

Choose the Best Starting Method

Use Prompt to Video when you have a topic but need help developing the complete structure.

Use Import Script when the wording is already finished and should remain unchanged.

Use a template when you want prepared layouts and a consistent visual style.

Use a blank video when you want full manual control.

Use Import PowerPoint when an existing presentation already contains the main content and visual structure.

Whichever method you choose, treat the first result as a draft. Review the script, remove unnecessary scenes, verify the information, and confirm that the selected layouts support the message before continuing with detailed editing.

Write and Organize the Video Script

The script controls what the AI presenter says, how long each scene lasts, and how the video’s information is divided. A clear script also makes it easier to choose suitable layouts, visuals, captions, and animations.

Add a Script to a Scene

1. Open the video in the Synthesia editor.
2. Select a scene from the list on the left.
3. Click the script box beneath the canvas.
4. Type or paste the words the presenter should say.
5. Select the language, voice, and speaker when needed.
6. Preview the narration before continuing.

Each scene has its own script. The wording can be edited at any point before the final video is generated.

Review the Complete Script

Synthesia’s expanded Storyboard view allows users to review the complete script without opening every scene individually.

1. Open the video.
2. Find the script box.
3. Click Expand in its upper-right corner.
4. Read and edit the full script.
5. Select Collapse to return to the regular canvas.

This view is useful for checking the overall structure, removing repetition, and confirming that the information flows logically between scenes.

Organize the Video Into Clear Sections

Begin with a simple structure:

1. Introduce the topic and explain why it matters.
2. Present the main information in a logical order.
3. Demonstrate or explain each important step.
4. Summarize the main points.
5. End with the action viewers should take next.

For example, a software tutorial might use separate sections for signing in, navigating the dashboard, completing the main task, and solving common problems.

Do not introduce every detail in the opening. Reach the main topic quickly, then explain supporting information when viewers need it.

Keep One Main Idea in Each Scene

Treat scenes like presentation slides. Each one should communicate one main idea, instruction, example, or transition.

A scene may become difficult to follow when it contains several unrelated points. Divide it when:

• The topic changes
• A new step begins
• A different visual is needed
• The presenter moves from explanation to demonstration
• The script becomes too long for one layout

Synthesia allows scenes to be added, duplicated, reordered, copied, and pasted through the scene menu on the left side of the editor. Scenes can also be copied between separate videos.

Use Short, Spoken Sentences

Write for listening rather than reading. Long sentences containing several clauses may sound unnatural when delivered by an AI voice.

For example:

Before:
“Synthesia provides several creation methods that users can choose from depending on whether they already have a completed script, an existing presentation, supporting documentation, or only a general idea for the video.”

Improved:
“Synthesia provides several ways to start a video. You can use a finished script, import a presentation, attach supporting documents, or begin with a simple idea.”

The revised version communicates the same information with clearer pauses and a more natural speaking rhythm.

Use Natural Transitions

Add brief transitions when moving between topics.

Useful examples include:

• “Now that the account is ready, let’s open the editor.”
• “Next, we’ll add an AI presenter.”
• “Before generating the video, review the captions.”
• “Let’s look at a practical example.”
• “Finally, check the finished export.”

Avoid using the same transition repeatedly. The script should guide viewers without announcing every small change.

Remove Unnecessary Wording

AI-generated scripts may include broad introductions, repeated conclusions, or phrases that do not add useful information.

Remove wording such as:

• “In today’s fast-paced digital world”
• “It is important to note that”
• “As we all know”
• “Without further ado”
• “In conclusion, to conclude”

Replace these phrases with the actual information the viewer needs.

Add Pauses

Pauses can make narration easier to understand and give viewers time to absorb an important statement. Synthesia’s script controls support pauses, and short pauses may also help expressive avatars deliver emotional or emphasized wording more naturally.

Use pauses after:

• Important statistics
• Questions
• Section introductions
• Complicated instructions
• Emotional or serious statements
• Lists containing several items

Do not add a long pause after every sentence. Too many pauses can make the narration feel slow and disconnected.

Check Names and Difficult Words

Preview names, abbreviations, technical terminology, product names, and words from another language before generating the complete video.

Synthesia’s pronunciation feature allows users to control how specific terms are spoken. Saved pronunciations can help maintain consistency when the same acronym, technical term, or brand name appears several times.

Correct pronunciation problems before adjusting scene timing because changing the wording may alter the narration’s length.

Add More Than One Speaker

A scene can include multiple avatars and speakers when the video requires a conversation, interview, role-play, or customer-service example.

To add another speaker:

1. Add the first avatar and write its line.
2. Select the correct language and voice.
3. Choose New Speaker from the script box.
4. Add the second avatar.
5. Enter the second speaker’s line.
6. Select that speaker’s voice.
7. Preview the conversation.

Synthesia connects each section of the script to the selected presenter and voice.

Keep conversations focused. A short exchange often communicates the example more clearly than a long discussion between several presenters.

Create Voiceover-Only Scenes

An avatar does not need to appear in every scene. Users can hide or remove the presenter while keeping the script and AI narration active. This creates a voiceover-only scene containing backgrounds, screen recordings, images, videos, or graphics.

Voiceover-only scenes work well for:

• Software demonstrations
• Product close-ups
• Charts and diagrams
• Full-screen presentations
• Step-by-step processes
• B-roll footage

Alternating between presenter scenes and full-screen visuals can create more variety than displaying the avatar continuously.

Upload Prerecorded Narration

Users can upload an audio recording instead of entering a typed script. The scene’s script box must be empty before the audio file is added, and the recording language must be selected after upload.

Uploaded narration is useful when a human voice, approved recording, or specific performance must be preserved. However, changing the spoken wording later usually requires editing or recording the audio again.

Match the Script to the Visuals

The narration should describe or support what appears on screen.

When the presenter says, “Select the Settings button in the upper-right corner,” the viewer should see that button or a clear screen recording of the action. Avoid showing generic office footage while giving precise software instructions.

Add a note while drafting when a scene requires a particular visual:

• Show the account-creation page.
• Display the three pricing options.
• Zoom in on the export button.
• Add a chart comparing the results.
• Use the presenter only during the introduction.

These notes can be removed from the spoken script after the scene has been designed.

Check the Estimated Pacing

Scene length is influenced by the script, narration, uploaded audio, animations, and media timing. Synthesia’s Timeline provides controls for adjusting when elements appear, while changes to the script or media can affect the scene’s overall duration.

When a scene feels too slow, shorten the script or add useful visual activity. When it feels rushed, simplify the wording, add a pause, or divide the information into separate scenes.

Complete a Final Script Review

Before designing every scene, confirm that:

• The introduction reaches the topic quickly.
• Each scene contains one main idea.
• The information appears in a logical order.
• Sentences sound natural when spoken aloud.
• Repeated and unnecessary wording has been removed.
• Names, numbers, and technical terms are accurate.
• Difficult words have been pronunciation-tested.
• Visual instructions match the narration.
• The conclusion provides a clear ending or next step.

A well-organized script makes the remaining editing process easier. Once the wording and scene structure are stable, the next step is selecting and customizing the AI presenter.

Add, Replace, and Customize AI Avatars

AI avatars act as the presenters in Synthesia videos. Users can choose a ready-made stock avatar, use a customizable synthetic presenter, or add a personal avatar created from their own photo or recording. The available avatar types and customization options depend on the account’s plan and the technology used to create the presenter.

Add an Avatar to a Scene

1. Open the project in the Synthesia editor.
2. Select the scene you want to edit.
3. Click Avatar in the top toolbar.
4. Browse or search the available presenters.
5. Select an avatar to add it to the canvas.
6. Drag the avatar into the preferred position.
7. Resize or adjust its appearance when necessary.
8. Enter the narration in the scene’s script box.

To remove the presenter, right-click the avatar and select Delete. Removing or hiding the avatar does not automatically remove the scene’s script or voiceover.

Choose the Right Avatar Type

Synthesia currently provides several avatar categories.

Express-1 stock avatars are professionally filmed presenters that support transparent backgrounds and Synthesia’s API. They may work well for long-form training, compliance material, and automated production, but they do not include generated hand gestures.

Express-2 stock avatars include more natural body movement, hand gestures, and flexible framing. Their original backgrounds are fixed and cannot be removed, making them better suited to conversational videos and product walkthroughs where movement is more important than a transparent presenter.

Synthetic stock avatars are completely AI-generated presenters rather than digital versions of real actors. They support customizable outfits, backgrounds, and avatar B-roll, making them one of the more flexible options for branded or frequently produced content.

Personal and Studio avatars use the appearance of a specific person. These are more useful when personal branding, leadership communication, or a recognizable presenter matters. Creating them will be covered later in this guide.

Replace an Existing Avatar

To change the presenter:

1. Select the avatar on the canvas.
2. Open the avatar dropdown in the editing panel.
3. Choose another presenter.
4. Review the replacement on the current scene.
5. Select Change All when the new avatar should replace every occurrence of the original presenter throughout the video.

Use Change All carefully when a project contains more than one speaker. Confirm that the selected avatar is connected to the correct script blocks so the narration and lip synchronization remain assigned to the intended presenter.

Reposition and Resize the Avatar

Select the avatar and drag it to another part of the canvas. Resize it so it remains visible without covering titles, captions, screenshots, products, or other important visual elements.

A larger presenter can work well for introductions and direct announcements. A smaller presenter may be more appropriate when the scene also contains a demonstration, diagram, slide, or substantial text.

Keep the presenter’s position reasonably consistent across related scenes. Sudden changes in size or placement can feel distracting unless they support an intentional layout change.

Hide the Avatar Without Removing the Voice

A presenter does not need to appear in every scene.

To create a voiceover-only scene:

1. Select the scene.
2. Click the eye icon beside the avatar in the script panel.
3. Keep the script and voice active.
4. Add a full-screen image, recording, slide, or other visual.
5. Preview the scene.

The avatar remains connected to the narration but is no longer visible. Avatar visibility is controlled separately for each scene, so one project can alternate between presenter-led sections and full-screen visual explanations.

Use an Avatar With a Space

A Space places the avatar inside a prepared environment, such as an office, studio, classroom, or other background.

To add one:

1. Select the avatar.
2. Find Space in the inspector panel.
3. Click Add Space.
4. Browse or search the available environments.
5. Select a background.
6. Adjust its appearance.

Spaces can be displayed full screen or inside circle and box layouts. Users can also adjust the crop, opacity, and shadow. Removing the Space through the canvas may remove the connected avatar as well, so confirm the result before continuing.

Choose an environment that supports the message. A classroom may suit a lesson, while an office or studio may appear more appropriate for business training. Avoid backgrounds containing unnecessary details that compete with the presenter.

Customize an Avatar’s Outfit and Environment

Synthesia allows supported customizable stock avatars and Personal Avatars created from photos to use generated outfits and spaces. These options are based on Express-2 avatar technology and may not appear for older or unsupported presenters.

To customize an avatar:

1. Open Avatar in the editor.
2. Select Customize Avatars or Create Outfit.
3. Choose the supported presenter.
4. Select a seated or standing pose when available.
5. Choose whether the avatar should appear on the left, center, or right.
6. Enter a prompt describing the outfit or environment.
7. Add available brand colors or a logo.
8. Generate the variations.
9. Review the results.
10. Select Use in Video.

Synthesia can also create a prompt based on the scene’s script. Generated variations generally take a few minutes and are saved to the user’s library after they are selected.

Write a Clear Outfit Prompt

Describe the clothing specifically enough for the result to match the subject.

For example:

“Dark green polo shirt with a simple white company logo on the chest, professional customer-support uniform.”

Users can include hexadecimal color codes when an exact brand color is needed. Logo placement is approximate, and highly detailed logos or small lettering may not appear accurately. A simplified icon usually works better than a complicated full logo.

Review the avatar’s face and skin tone after generating a new outfit. Synthesia notes that customization may cause small color differences, so another variation may be needed to achieve a closer match.

Adjust the Pose and Framing

Supported customizable avatars can be generated in seated or standing positions and placed toward the left, center, or right side of the environment. Their pose and zoom can also be changed through the camera controls.

Choose framing based on what else appears in the scene. Position the presenter on one side when text or media must appear beside them. A centered presenter works better when the scene focuses primarily on the speaker.

Do not place the avatar too close to the edges. Important parts of the presenter may be cropped when the video is viewed on a different screen or adapted to another aspect ratio.

Add Multiple Avatars

Synthesia supports dialogue scenes containing multiple presenters. Users can add as many as 20 avatars to one scene and assign a separate voice and script block to each one.

To create a simple dialogue:

1. Add the first avatar.
2. Write the first speaker’s script.
3. Choose that speaker’s voice and language.
4. Add another avatar from the Avatar menu or select New Speaker.
5. Write the second speaker’s line.
6. Assign the correct voice.
7. Position both presenters clearly.
8. Preview the conversation.

Dialogue scenes can be useful for interviews, role-playing exercises, customer-service examples, and training scenarios. Keep each speaker’s lines brief so viewers can follow the exchange easily.

Match Each Avatar to the Correct Script

When several presenters appear in one scene, confirm that each script block is assigned to the correct avatar and voice. An incorrect assignment may cause the wrong presenter to move or create lip-synchronization problems.

Speakers can also remain hidden while their voices play. This may be useful for an off-screen question, background narrator, or conversation where only one participant needs to appear at a time.

Understand the Editor Preview

Avatars do not fully animate or move their lips while the project is being edited. Synthesia keeps the editor preview simplified so scenes can load more quickly. The complete animation and lip synchronization appear after the video is generated.

Because the editor does not show the complete performance, preview the selected voice carefully and check the generated video afterward. Small adjustments to the script, punctuation, pauses, and voice speed can affect how natural the final avatar appears.

Use Avatars Selectively

Displaying the same presenter throughout every scene may make a video feel repetitive. Consider using the avatar during:

• The introduction
• Important explanations
• Transitions between sections
• Direct calls to action
• The conclusion

Hide the avatar during detailed software demonstrations, charts, product close-ups, and other scenes where viewers need to focus on the full-screen visual.

Review the Avatar Before Continuing

Confirm that the chosen presenter fits the intended audience, topic, and tone. Check that the avatar does not cover other content, the outfit and background feel appropriate, and the correct script and voice are assigned.

Also avoid scripting a stock avatar as though it is the real actor sharing personal experiences, beliefs, or endorsements. Synthesia requires avatar use to follow its content-moderation and impersonation rules.

Once the presenter and scene placement are correct, the next step is choosing an AI voice and correcting pronunciation.

Choose an AI Voice and Correct Pronunciation

Synthesia automatically assigns a voice to an avatar, but users can change the language, regional accent, voice, and speaking speed. The selected voice should match the intended audience and sound natural with the script.

Change the Voice

1. Select the scene you want to edit.
2. Click the speaker pill inside the script box.
3. Open the Voice dropdown.
4. Browse the recommended voices.
5. Use the search and filter controls when needed.
6. Preview a voice before selecting it.
7. Apply the voice to the current scene.
8. Select Change All when the same replacement should be used throughout the video.

Synthesia’s stock voice library covers more than 140 languages and accents. Available voices can be filtered by language, accent, or gender, although the exact selection may vary by plan and avatar.

Choose the Correct Language and Accent

Synthesia usually detects the script’s language automatically. Open the voice-selection menu when it chooses the wrong language or when the video requires a particular regional accent.

For example, English may include US, UK, Australian, and other regional options. Spanish may offer separate voices for Spain, Mexico, the United States, and other supported regions.

Choose the variation that matches the audience. A training video for employees in Spain may sound more appropriate with a Spanish voice from Spain than one designed for a Latin American audience.

Not every voice is offered in every dialect. A presenter or voice available for one regional version of a language may be unavailable in another.

Preview Several Voices

Do not select a narrator based only on its label. Listen to several samples and compare how they handle the actual script.

Synthesia’s Voices page allows users to browse voices by language, accent, and gender. Users can also enter their own text to create and download a sample audio clip.

Test the voice with a sentence containing:

• The company or product name
• Important technical terms
• Numbers and abbreviations
• Questions
• Emotional or emphasized wording

A voice that sounds natural in a short demonstration may behave differently with the real script.

Match the Voice to the Content

Choose a narrator based on the topic and presentation style rather than simply selecting the most energetic option.

A calm and clear voice may work well for compliance training, tutorials, and detailed explanations. A conversational voice may feel more natural for onboarding or customer education. A confident and energetic voice may suit announcements, promotional content, or short explainers.

Keep the same primary voice across related scenes unless the video includes multiple speakers. Changing narrators without a clear reason can make the project feel inconsistent.

Adjust the Speaking Speed

Synthesia allows supported AI voices to be adjusted from approximately 0.8x to 1.2x speed.

1. Hover over the speaker pill.
2. Select the speed control.
3. Move the slider toward a slower or faster setting.
4. Preview the narration.
5. Select Change All when the adjustment should apply to every script block using that avatar and voice.

Speed controls currently apply to supported voices provided through Synthesia and certain integrated voice providers. The displayed speed is approximate, so a 1.2x setting may not make every sentence exactly 20 percent faster.

Slow down sections containing complicated instructions, unfamiliar terminology, or important safety information. A slightly faster speed may work for simple transitions or repeated information.

Avoid using speed controls to force an excessively long script into a short scene. Shortening the script or dividing it into additional scenes usually sounds more natural.

Add Natural Pauses

Pauses give viewers time to understand important information and can make the AI narration sound less mechanical.

Place the cursor where the pause should occur, then use the script controls or right-click menu to add it. Synthesia allows users to define the pause duration.

Useful places for pauses include:

• After a question
• Before an important answer
• Between steps
• After a statistic
• Before changing topics
• Before the final call to action

Synthesia recommends small pauses when working with expressive avatars because they can improve the delivery and give emotional statements more natural spacing.

Use pauses selectively. Adding one after every sentence can make the narrator sound slow and disconnected.

Correct a Mispronounced Word

Synthesia’s pronunciation tool can change how acronyms, technical terms, brand names, and other difficult words are spoken.

1. Highlight the word or phrase in the script.
2. Select Pronunciation.
3. Choose Type to enter a phonetic spelling or Record to speak the correct pronunciation.
4. Listen to the automatic preview.
5. Adjust the pronunciation when necessary.
6. Select Use to save it.

For example, a brand name written as “Porsche” could be assigned a pronunciation that guides the narrator toward “Porsh-uh.”

The written caption can continue displaying the correct spelling while the voice uses the pronunciation instruction.

Use Phonetic Spelling Carefully

When the pronunciation control does not produce the desired result, rewrite the word according to how it should sound.

For example:

• Write “A.I.” when “AI” is read as one word.
• Write a number in words when the numerical version sounds unnatural.
• Divide an abbreviation with spaces or punctuation.
• Spell a foreign name phonetically.
• Add commas to create shorter pauses.

For abbreviations in a non-English script, Synthesia recommends spelling the term according to how it should sound in that language.

Do not change visible on-screen text to an incorrect spelling when viewers need to see the official term. Use the pronunciation feature whenever possible.

Save Repeated Pronunciations

Pronunciations that will be reused can be stored in the workspace glossary. Once saved, a glossary pronunciation can be applied consistently across videos using that language rather than being corrected separately in every project.

This is particularly useful for:

• Company names
• Product names
• Employee names
• Industry terminology
• Abbreviations
• Locations
• Frequently mentioned foreign words

Review glossary entries when switching voices. A pronunciation that sounds correct with one narrator may need adjustment with another.

Correct Questions and Emphasis

End questions with a question mark. Without the correct punctuation, the AI voice may deliver the line like a statement.

When the result still sounds unnatural, regenerate the audio or rewrite the question. Synthesia notes that voice generation can vary, so another generation may produce a better delivery.

Punctuation can also guide emphasis and pacing. Commas, periods, shorter sentences, and carefully placed pauses often sound more natural than using one long sentence.

For expressive avatars, the wording and punctuation can affect both vocal delivery and automatically generated gestures. Those gestures cannot always be disabled individually, so use a non-expressive avatar when more controlled movement is required.

Use Multiple Languages in One Scene

One avatar can speak more than one language within a scene, but each language or voice change must begin on a separate line. Synthesia cannot switch voices in the middle of one sentence.

To use several languages:

1. Enter the first language on one line.
2. Begin the second language on a new line.
3. Check whether Synthesia detects it automatically.
4. Open the speaker pill when the wrong voice is assigned.
5. Select the correct language and regional voice.
6. Preview each line separately.

Use multilingual scenes carefully. Frequent changes may become difficult to follow, and some voices may not be available across every required dialect.

Use a Custom Voice Clone

Synthesia allows users on all plans to create a voice clone by recording inside the platform or uploading an existing audio sample. Some supported languages remain limited to Enterprise users.

After creation, the clone appears under the custom voices in the voice picker. The avatar and voice are selected separately, so a custom voice can be paired with different available avatars.

Voice cloning requires the actual speaker to provide recorded consent. When uploading audio, Synthesia currently accepts recordings between one and five minutes in several common audio formats.

The complete voice-cloning process will be covered in the next section.

Understand Voice Availability

Voice options may change over time because Synthesia works with several voice technologies and external providers. Previously generated videos continue to work when a voice is retired, but editing or regenerating the project may cause Synthesia to substitute another available voice.

Before producing a large series, test the preferred voice and confirm that it is available under the correct workspace and plan. For long-term consistency, consider using an approved custom voice clone when appropriate.

Review the Narration Before Continuing

Listen to the full script rather than previewing only the first scene. Confirm that:

• The language and regional accent are appropriate.
• The narrator matches the video’s tone.
• Names and technical terms are pronounced correctly.
• Questions sound like questions.
• Pauses appear in natural locations.
• The speed remains comfortable.
• Volume and vocal style remain consistent.
• Multiple speakers use the correct voices.

Correct the narration before completing detailed animations and scene timing. Changes to pronunciation, punctuation, pauses, or speed may affect the duration of a scene and the timing of its visual elements.

Create a Personal Avatar and Cloned Voice

A Personal Avatar allows users to create Synthesia videos with their own appearance instead of using a stock presenter. Synthesia currently supports Personal Avatars created from a single photo or from recorded video footage. Personal Avatars are available on Starter plans and above.

Choose a Creation Method

The two main options are:

Express-2 Avatar from a photo uses one clear image to create a presenter with generated body language. This method is faster, supports customizable clothing and environments, and does not require a full training video.

Legacy Avatar from video learns from a recorded or uploaded video of the user. It can remove the original background and automatically creates a voice clone from the submitted recording. Processing typically takes about one business day.

A photo-based avatar is usually the easier option for beginners. Use the video-based method when you prefer a presenter created from recorded footage or need its supported background-removal workflow.

Create a Personal Avatar From a Photo

1. Open Avatars from the left-hand menu.
2. Select Create Avatar.
3. Choose Create Your Personal Avatar.
4. Select Create My Avatar.
5. Choose Express-2 Avatar.
6. Upload a photo or take one with the webcam.
7. Record an optional voice sample.
8. Record the required consent video.
9. Submit the avatar for processing.
10. Wait for it to appear in the Avatars library.

Synthesia states that photo-based avatars normally generate within minutes, although identity or consent submissions requiring manual review may take longer.

Choose a Suitable Photo

Use a clear, well-lit photograph containing only the person whose avatar is being created. A waist-up image works well because it provides enough information for the presenter’s upper body and face.

Synthesia recommends showing the person’s teeth because this can improve lip synchronization. The quality of the photograph directly affects how realistic the generated avatar looks and moves.

Avoid using:

• Blurry or poorly lit photos
• Images containing several people
• Heavy filters or facial alterations
• AI-generated portraits
• Illustrations, mannequins, or fictional characters
• Photos of celebrities or other people

The uploaded image must show a real person and match the person who completes the live consent recording.

Understand Photo-Avatar Movement

A photo-based avatar does not learn the user’s personal gestures or body language from one still image. Instead, Synthesia generates suitable movements based on the spoken script.

The avatar may therefore resemble the user without moving exactly as they would in real life. Small differences in facial details, expressions, proportions, or appearance may also occur.

Create a brief test video before using the avatar in a large project. Check its facial appearance, lip synchronization, posture, gestures, and framing.

Customize a Photo-Based Avatar

Once the avatar has been created, users can customize its outfit and environment or generate action-based avatar B-roll.

For example, the same presenter could appear:

• In professional office clothing
• Wearing a branded uniform
• Sitting in a studio
• Standing in a classroom
• Demonstrating an activity in B-roll footage

These generated variations should still be reviewed closely. Logos, clothing details, colors, and facial features may not remain perfectly consistent across every generation.

Create a Personal Avatar From Video

To use recorded footage:

1. Open Avatars.
2. Select Create Avatar.
3. Choose Create Your Personal Avatar.
4. Select Create My Avatar.
5. Choose Legacy Avatar.
6. Select whether to record with a webcam or upload existing footage.
7. Review the recording requirements.
8. Record or upload the video.
9. Choose whether to remove the background.
10. Record the separate consent video.
11. Confirm the biometric-consent statement.
12. Submit the files for processing.

The person in the main footage and consent recording must be the same. Submissions can be rejected when Synthesia cannot verify the identity or permission clearly.

Record With a Webcam

When recording inside Synthesia:

1. Select the correct camera and microphone.
2. Choose the recording language.
3. Follow the displayed script and instructions.
4. Record the complete performance.
5. Review the footage.
6. Record again when necessary.
7. Choose whether to remove the background.
8. Complete the separate consent recording.

Synthesia processes video-based Personal Avatars in approximately one business day under normal conditions.

Upload Existing Footage

Uploaded footage should:

• Be one continuous take
• Be between one and five minutes long
• Contain only one person in the foreground
• Use MP4, MOV, or WebM format
• Remain under 2 GB

Users can enable background removal during submission. The footage does not have to be recorded directly from the front, although clear framing and visibility generally produce a stronger result.

Record High-Quality Avatar Footage

Use a quiet, evenly lit space with a stable camera. Position the camera close to eye level and keep the person clearly separated from the background.

During recording:

• Look toward the camera.
• Speak naturally and clearly.
• Keep your hands away from your face.
• Avoid sudden movements.
• Do not cut or combine multiple takes.
• Keep other people out of the foreground.
• Avoid loud background noise.

The avatar is generated from the submitted footage, so unclear audio, low lighting, excessive movement, or poor framing may affect the finished result.

Complete the Consent Recording

Synthesia requires explicit consent before creating an avatar based on a real person. The consent recording must be completed by the same person shown in the photo or video.

For a photo avatar, the user records a live video confirming that they understand their likeness will be used to create a digital presenter. They must also read the displayed passcode aloud. Edited, replayed, prerecorded, or unclear consent footage may be rejected.

A user cannot create a photo-based avatar on another person’s behalf, even when that person has informally provided permission. The person whose likeness is used must complete the required creation and consent process directly.

Request an Avatar From a Colleague

Organizations can send an avatar-creation request when an employee, client, instructor, or other participant needs to create their own presenter.

1. Open Avatars.
2. Select Create Avatar.
3. Choose Create Your Personal Avatar.
4. Select Request From Colleague.
5. Choose a photo-based or video-based avatar.
6. Enter the recipient’s name and email address.
7. Add an optional explanation.
8. Send the request or copy the invitation link.

The recipient does not need an existing Synthesia account. The invitation opens a separate creation page where they submit their own image or recording and complete the consent requirements. Request links currently expire seven days after creation.

Once the process is complete, the requested avatar is owned by the person or workspace that sent the request. Both parties receive notifications as the avatar is processed.

Create a Voice Clone With the Avatar

A voice clone is automatically created when users make a Personal Avatar from video footage. Photo-avatar users can optionally record a voice sample during creation. When this step is skipped, the avatar can use one of Synthesia’s existing AI voices instead.

The cloned voice and Personal Avatar remain separate assets. The Personal Avatar can use another Synthesia voice, and the cloned voice can be paired with another available avatar.

Create a Voice Clone Separately

Voice cloning is currently available across all Synthesia plans, although some supported languages are limited to Enterprise accounts. Users can create a voice by recording directly in Synthesia or uploading an existing audio file.

To create one:

1. Open the Voices area.
2. Select Create Voice.
3. Choose Record Your Voice or Upload an Audio File.
4. Enter the requested language and speaker information.
5. Record or upload the voice sample.
6. Complete the consent recording.
7. Read the randomly generated passcode aloud.
8. Submit the voice for generation.
9. Wait for it to appear under Custom Voices.

Record a Strong Voice Sample

When recording inside Synthesia, use a good microphone in a quiet room. Read the displayed script with a positive, natural tone, pause between paragraphs, and take normal breaths.

The speaker must then read a randomly generated consent passcode aloud. The consent recording should be under 60 seconds and cannot be silent.

Keep the delivery consistent. Large changes in volume, distance from the microphone, energy, or accent can make the generated voice less stable.

Upload a Voice Sample

Uploaded audio must be between one and five minutes long. Synthesia currently supports MP3, WAV, M4A, WEBA, AAC, FLAC, and OGG files for this workflow.

Use an audio file that contains:

• One speaker
• Clear speech
• Minimal echo
• No background music
• No overlapping voices
• Consistent microphone quality
• Natural pacing and expression

The actual speaker must complete the consent step. Consent cannot be provided by another person on the speaker’s behalf.

Use the Voice Clone in a Video

1. Open the project.
2. Select the correct scene.
3. Open the script box.
4. Open the avatar and voice controls.
5. Select the presenter.
6. Open the full voice list.
7. Choose the cloned voice under Custom Voices.
8. Preview the script.
9. Apply it to additional scenes when needed.

Because the voice and avatar are selected separately, choosing a Personal Avatar does not always apply its cloned voice automatically. Confirm both selections for every speaker.

Test the Voice in Other Languages

Synthesia allows a cloned voice to be used in multiple supported languages, not only the language used for the original recording.

However, pronunciation, rhythm, and accent may not remain equally natural in every language. Create a short sample and ask a fluent speaker to review it before producing a complete translated video.

Synthesia also notes that its instant voice-cloning technology may not perfectly preserve every non-native English accent.

Protect Avatar and Voice Access

Custom avatars and voice clones remain private to their creator and are not added to Synthesia’s public stock library. Sharing them with team members is currently an Enterprise feature.

Only the creator can authorize sharing. Custom assets can generally be shared only within the same workspace unless Synthesia support completes an authorized transfer.

Organizations should decide who can use a personal avatar or voice, what types of videos are permitted, and what should happen to those assets if the person leaves the company.

Delete an Avatar When It Is No Longer Needed

To delete a Personal Avatar:

1. Open Avatars.
2. Locate the presenter.
3. Open its three-dot menu.
4. Select Delete.
5. Confirm the deletion.

Deleting a video-based Personal Avatar returns its avatar allowance to the account. Review existing projects before deleting it because projects using that presenter may need another avatar assigned before they can be regenerated.

Use Personal Avatars Responsibly

Create an avatar or voice only for yourself or a person who has knowingly completed Synthesia’s consent process. Do not use someone’s photograph, footage, or recording to imitate them without direct authorization.

A Personal Avatar is most useful when a recognizable presenter adds trust or consistency, such as in leadership updates, employee training, customer education, personal branding, or multilingual communication.

Before using it in a major project, generate a brief sample and review the likeness, voice, gestures, lip synchronization, pronunciation, and visual quality. Correcting problems during a short test is easier than regenerating a long finished video.

Add Scenes, Layouts, Backgrounds, Images, Video, and Text

Scenes are the main building blocks of a Synthesia video. Each scene works like a presentation slide and can contain an avatar, script, text, images, video clips, shapes, and other visual elements. Keeping one main message in each scene makes the finished video easier to follow.

Add a New Scene

1. Open the video in the editor.
2. Select the scene that should come before the new one.
3. Click Add Scene in the canvas toolbar.
4. Choose a blank scene or a prepared template layout.
5. Add the script and visual elements.
6. Preview the new scene with the surrounding content.

The new scene appears directly after the scene that was selected. A blank scene provides full control, while a template gives users a prepared arrangement for text, media, and avatars.

Use a new scene when the topic changes, another step begins, a different visual is required, or the current scene contains too much information.

Duplicate or Copy a Scene

Duplicating a completed scene can save time when the next section should use the same design. After creating the copy, replace its script, images, and other content while preserving the original layout.

Scenes can also be copied from one Synthesia video and pasted into another. The copied version includes its script, visual and audio elements, and animations. Animation timing may need adjustment when the source and destination scenes have different durations.

This is useful for reusing:

• Branded introductions
• Section-title scenes
• Product layouts
• Speaker introductions
• Calls to action
• Closing scenes

Rename and review the copied content so no text, images, or narration from the previous project remain accidentally.

Reorder or Delete Scenes

Drag scenes into a different position when the information appears in the wrong order. After moving one, read the scripts before and after it to make sure the transitions still make sense.

To remove a scene, right-click it and choose Delete, or select it and press the Delete key. Deleting the scene also removes its connected script and visual content from the project.

Before deleting, confirm that its information is unnecessary rather than simply presented poorly. Replacing the layout or shortening the script may solve the problem without removing the content.

Replace a Scene Layout

Layouts control where avatars, titles, media, and other elements appear.

To replace one:

1. Find the scene in the scene list.
2. Right-click the scene.
3. Select Replace Layout.
4. Browse the available options.
5. Choose a new layout.
6. Review every element on the canvas.

Synthesia preserves the scene’s script when a layout is replaced, but some titles, subtitles, images, or other visual elements may shift or disappear. Check the scene carefully after applying the new design.

Choose layouts based on what the viewer needs to see. A presenter-and-text layout may suit an introduction, while a full-screen media layout is usually better for software demonstrations or product footage.

Use Layouts Consistently

A video does not need the same layout in every scene, but its design should still feel connected.

A simple structure might use:

1. A title layout for the opening.
2. A presenter layout for explanations.
3. A media-focused layout for demonstrations.
4. A list or comparison layout for key points.
5. A summary layout for the conclusion.

Avoid changing the entire design in every scene simply to create variety. Consistent spacing, fonts, colors, and avatar placement help viewers focus on the information.

Add an Image or Video

To add media:

1. Select the scene.
2. Click Media at the top of the canvas.
3. Choose Upload Media or browse the available stock libraries.
4. Select an image or video.
5. Move, resize, and position it on the canvas.
6. Use the inspector controls to adjust its appearance.
7. Preview the scene.

Synthesia currently accepts JPEG, PNG, and SVG images up to 300 MB. Supported uploaded video formats include MP4 and WebM files up to 250 MB.

Use original media when accuracy matters, particularly for products, software interfaces, employees, workplaces, charts, and company procedures.

Browse Stock Media

Synthesia includes integrated stock-image and stock-video libraries. Basic-plan users may see watermarks on stock previews, while paid plans receive access to full-resolution Getty media without those preview watermarks.

Search with specific descriptions. For example:

• “Customer-support employee wearing a headset”
• “Warehouse worker scanning a package”
• “Small-business owner reviewing online sales”
• “Cybersecurity dashboard on a laptop”

A broad search such as “work” may return visually attractive footage that does not represent the script accurately.

Use Media as the Background

Any supported image or video can be used as a full-scene background.

1. Select the scene.
2. Turn on Background Media in the right-side toolbar.
3. Choose a solid color, stock visual, or uploaded file.
4. Review how the avatar and text appear over it.
5. Adjust the contrast, positioning, or blur when necessary.

The selected visual automatically fills the scene background.

When using a busy background, add a dark or light overlay behind the text so it remains readable. Avoid placing important wording directly over faces, bright windows, detailed charts, or moving objects.

Resize and Crop Media

Select an image or video and drag its handles to resize it. Reposition the asset until its most important subject remains visible.

Check the framing whenever:

• A landscape image is placed in a portrait project.
• A vertical video is added to a widescreen scene.
• A person is positioned near the edge.
• Text must appear beside the media.
• The visual contains its own written information.

When an uploaded image appears unexpectedly blurry, Synthesia recommends confirming that its Blur setting is at 0% in the inspector panel.

Store Reusable Media

The Media Library stores assets that need to be used across several projects. Media saved there is available to users within the same workspace. Supported library assets include images, MP4 and WebM videos, GIFs, and Lottie JSON animations.

To upload an asset:

1. Open Library under the Assets section.
2. Select Media Library.
3. Click Upload Media.
4. Select the file.
5. Rename it clearly after uploading.

The three-dot menu can be used to rename, download, or delete stored media. Synthesia does not currently support folders inside the Media Library, so clear asset names are especially important.

Generate an AI Image or Video

Synthesia can also create custom images and video clips from text prompts through the Create With AI area.

1. Open Media.
2. Select Create With AI.
3. Choose Image or Video.
4. Describe the visual.
5. Select an available generation model.
6. Review the credit cost.
7. Generate the asset.
8. Add the result to the scene.

Users can press Tab to let Synthesia create a suggested prompt based on the video’s script. Generated assets remain available at the individual user level and can be added to other scenes or projects. This feature is offered as an add-on and uses credits, although some Enterprise image-generation access has different terms.

Describe the subject, action, environment, visual style, camera framing, and lighting.

For example:

“Customer-support specialist answering a video call in a modern home office, realistic corporate photography, natural lighting, medium shot.”

Review generated visuals for distorted faces, unusual hands, incorrect text, changing products, or unrealistic motion before using them.

Add Text to a Scene

To add on-screen text:

1. Select Text at the top of the canvas.
2. Choose Title 1, Title 2, Subtitle, Body, or Caption.
3. Click inside the new text box.
4. Replace the placeholder wording.
5. Drag the box into position.
6. Resize it using the corner handles.

Dragging a corner changes the text size dynamically, while dragging the box itself changes its location on the canvas.

Use titles for section names, body text for brief explanations, and captions or labels for supporting information. Avoid placing the full spoken script on screen unless viewers genuinely need to read it.

Style the Text

Select a text element to open its formatting controls. Synthesia supports bold, italics, underlining, font-family changes, text sizing, colors, line-height adjustments, letter spacing, and repositioning within the text box.

A readable text style should have:

• Strong contrast with the background
• A clear font
• Enough size for mobile viewing
• Comfortable line spacing
• Short paragraphs
• Consistent alignment

Custom font uploads are currently limited to Enterprise plans. Supported upload formats include TTF, OTF, WOFF, and WOFF2. Once added, those fonts become available throughout the workspace.

Keep On-Screen Text Concise

Viewers should not have to choose between reading a long paragraph and listening to the presenter.

Use on-screen text for:

• Section titles
• Important terms
• Short instructions
• Statistics
• Names and job titles
• Brief summaries
• Calls to action

When a scene needs several sentences of written information, divide it into shorter sections or use a voiceover-only scene so the text can receive more space.

Arrange Several Elements

Scenes may contain an avatar, title, image, logo, and several additional design elements. Select multiple assets by holding Shift while clicking them. The selected elements can then be grouped and moved together.

Lock completed elements to prevent accidental movement. Synthesia’s Tidy control can also automatically align and distribute selected elements when the layout looks uneven.

Use grouping for elements that should always remain together, such as:

• An icon and its label
• A product image and price
• A speaker photo and name
• A chart and its title

Control Which Element Appears in Front

When assets overlap, their layer order determines which one is visible. The Timeline allows users to drag layers up or down, rename them, group them, and control when they appear.

For example, a dark overlay should appear above the background but below the text. A logo may need to remain above both the background and another image.

Rename layers in complicated scenes so items such as “Product Image,” “Headline,” and “Avatar” are easier to identify.

Animate Visual Elements

Images, text, avatars, and other supported elements can enter or exit at specific points in the narration.

1. Select the element.
2. Find its entry or exit trigger in the script or Timeline.
3. Drag the trigger to the word where the animation should begin.
4. Choose an animation style.
5. Adjust its direction, speed, duration, or delay.
6. Preview the scene.

Available styles currently include Instant, Fade, Move, Slide, Scale, and Zoom. Synthesia uses Instant as the default.

Animations should clarify the information rather than decorate every object. For example, display each numbered step as the presenter explains it instead of showing the full list immediately.

Use the Timeline for Detailed Timing

Open the Timeline tab when a scene contains several elements that must appear in a particular order. The Timeline displays the visual layers and allows users to reposition animations, apply delays, organize groups, and adjust animation duration.

The scene’s duration is determined by its longest element. A video clip or animation that continues beyond the narration can therefore keep the scene playing after the presenter has finished speaking.

When a scene ends too slowly, check for an unnecessarily long video, delayed animation, or visual element that remains active after the narration.

Review the Complete Scene

Before continuing, preview each scene and confirm that:

• The layout supports the main idea.
• The avatar does not cover important content.
• Images and videos match the narration.
• Text remains readable.
• Media is correctly cropped.
• Visual elements are aligned.
• Animations appear at the right moment.
• The scene does not continue after the message is complete.

Use visual variety without overcrowding the canvas. The strongest scenes normally combine a clear message with only the elements needed to explain it.

Record Your Screen and Add Media

Synthesia’s AI Screen Recorder can capture a browser tab, application window, or full screen and turn the recording into editable video scenes. It is useful for software tutorials, product demonstrations, onboarding videos, and step-by-step instructions.

When microphone audio is included, Synthesia can transcribe the narration, remove filler words, and connect the recording to an editable script.

Install the Screen Recorder

The recorder is available through a browser extension for Google Chrome and Microsoft Edge.

1. Open or create a Synthesia video.
2. Select Record above the canvas.
3. Choose Install Extension.
4. Continue to the browser’s extension store.
5. Add the Synthesia extension.
6. Approve the requested permissions.
7. Return to the Synthesia editor.

The extension allows the recording controls to remain available while users move between browser tabs.

Start a Screen Recording From the Editor

1. Open the video where the recording should appear.
2. Select Record.
3. Choose Record Screen.
4. Select Full Screen, Window, or Browser Tab.
5. Choose a microphone or select No Microphone.
6. Click Start Recording.
7. Select the content to capture.
8. Click Share.
9. Wait for the three-second countdown.
10. Complete the demonstration.
11. Select the stop control when finished.

Synthesia automatically adds the recording to a new scene after processing it. When microphone narration was recorded, the platform can also transcribe the speech and divide longer content into logical scenes.

Choose What to Capture

Use Browser Tab when the demonstration takes place entirely inside one website. This option also supports Synthesia’s blur tool for hiding sensitive information.

Use Window when recording one application, such as spreadsheet software, presentation software, or a desktop program.

Use Full Screen when the demonstration requires switching between several applications or windows. Be careful because notifications, personal files, and unrelated windows may appear in the recording.

Close unnecessary applications and remove private information before beginning.

Record With or Without a Microphone

Choose a microphone when you want to explain the process naturally while performing it. Synthesia converts the spoken narration into editable script text after the recording.

Choose No Microphone when you prefer to add a polished AI voiceover afterward. This can be easier when the demonstration requires concentration or when the script needs to be approved before narration is generated.

A practical beginner process is to record the actions without speaking, then add a clear script scene by scene. This separates the visual demonstration from the narration and makes mistakes easier to correct.

Prepare Before Recording

Practice the process once before capturing it. Confirm that all pages, files, accounts, and examples are ready.

Before recording:

1. Close unnecessary tabs and applications.
2. Turn off desktop notifications.
3. Remove private customer or account information.
4. Increase the interface size when buttons are difficult to see.
5. Place the cursor near the first action.
6. Write a short list of the steps.
7. Perform a brief test recording.

Move the cursor deliberately and pause briefly after important clicks. Extremely fast movements can make the demonstration difficult to follow.

Blur Sensitive Information

The browser extension’s blur tool can conceal passwords, email addresses, customer records, account numbers, or other private information.

To add blur before recording:

1. Open the browser tab you will capture.
2. Select the Synthesia extension.
3. Click Blur.
4. Hover over the areas that should be hidden.
5. Apply the blur.
6. Select Done.
7. Start recording.

Blur can also be added during a recording. Opening the blur control pauses the capture while users select additional areas. The recording can then be resumed.

The blur feature currently works when capturing a browser tab, not when recording an entire screen or application window.

Whenever possible, replace real personal information with demonstration data instead of relying entirely on blur.

Record Through the Browser Extension

A screen recording can also be started without first opening a video.

1. Select the Synthesia extension in the browser toolbar.
2. Confirm the correct workspace.
3. Choose the current tab, a window, or the full screen.
4. Select the microphone.
5. Add blur when necessary.
6. Begin the recording.
7. Complete the demonstration.
8. Stop the recording.
9. Allow Synthesia to open the new project.

Recordings started directly from the extension are opened as new Synthesia videos. Synthesia applies a default template in this workflow rather than automatically using a previously selected template or Brand Kit.

Review the design afterward and apply the correct branding manually.

Upload an Existing Screen Recording

Users can import an MP4 recording created with another screen-capture program.

1. Open or create a video.
2. Select Record.
3. Choose Upload Screen Recording.
4. Select the MP4 file.
5. Wait for Synthesia to process it.
6. Confirm the recording language.
7. Select None when the recording has no spoken narration.
8. Review the generated script and scenes.

Synthesia can transcribe uploaded narration, remove filler words, and divide recordings longer than five minutes into separate scenes. Uploaded screen recordings can also use the project’s existing template and Brand Kit.

Review the Generated Transcript

After processing, read the transcript while watching the recording. Correct names, technical terms, button labels, and any words Synthesia misunderstood.

Removing filler words can make the narration more concise, but automatic edits may occasionally create an unnatural sentence. Listen to the sections around each edit and restore wording when necessary.

Check that:

• The transcript matches what was actually said.
• Instructions remain in the correct order.
• Important warnings were not removed.
• Product names and menu labels are spelled correctly.
• Scene breaks occur at logical moments.

The transcript becomes part of the project’s editable script, so correcting it early makes captions and later narration easier to manage.

Trim the Recording

Remove preparation time, mistakes, loading delays, or unnecessary footage from the beginning and end.

1. Select the recording on the canvas.
2. Choose Trim from the video controls.
3. Drag the beginning or ending handles.
4. Use Split at Time when a section in the middle needs separate control.
5. Preview the selected portion.
6. Click Done.

Trimmed footage is not permanently removed. The handles can be expanded later to restore previously hidden sections.

Avoid trimming so closely that the recording begins in the middle of a cursor movement or ends immediately after a click. Leave enough time for viewers to understand the action.

Split a Long Recording

A long recording may be easier to edit when divided into shorter scenes. Separate it when:

• A new task begins
• The software page changes
• The narration moves to another topic
• A pause or explanation is needed
• The viewer needs to focus on a particular button or setting

Shorter scenes also make it easier to add titles, avatars, text, zoomed screenshots, and different narration.

Synthesia may automatically divide recordings longer than five minutes, but users should still review whether the generated breaks occur at natural points.

Match the Recording to the Scene Duration

Select the video asset and choose how it should behave when its duration differs from the scene.

Play Once plays the clip one time and leaves the remaining scene duration unchanged.

Stretch adjusts the playback speed so the clip fills the scene.

Loop/Cut repeats a short clip or cuts a longer one when the scene ends.

These playback controls can be combined with trimming. For example, users can isolate the most useful portion of a recording and then loop or stretch it to match the narration.

Avoid stretching instructional recordings excessively. Speeding up a software demonstration may make clicks and menu changes difficult to follow.

Add an Avatar Beside the Recording

An avatar can introduce the task, explain important steps, or summarize what viewers should notice.

A useful structure is:

1. Show the avatar during the introduction.
2. Switch to the full-screen recording for detailed actions.
3. Bring the avatar back for warnings or explanations.
4. End with the avatar summarizing the next step.

When the recording contains small interface elements, hide the avatar so the demonstration can use the complete canvas.

Add Supporting Text

Use brief text to identify buttons, keyboard shortcuts, warnings, or important steps. For example:

• Select Account Settings
• Turn on Email Notifications
• Do not close this window
• Press Ctrl + S
• Processing may take several minutes

Do not cover the area where the action occurs. Place text beside the interface or use an arrow, shape, or highlight to direct attention.

Add Uploaded and Stock Media

Additional images, videos, graphics, and audio can be added through Media above the canvas.

1. Select the scene.
2. Open Media.
3. Upload a file or browse the available stock libraries.
4. Select the asset.
5. Move and resize it on the canvas.
6. Adjust its timing when necessary.

Uploaded and stock media can support the screen recording with product images, diagrams, close-up screenshots, or examples that are difficult to show during the live demonstration.

Keep supporting media relevant. A tutorial usually benefits more from an accurate screenshot than from generic office footage.

Reuse Media Across Projects

Frequently used images and videos can be stored in the Media Library. Assets saved there are available to other users in the same workspace.

Synthesia’s Media Library currently supports JPG, PNG, and SVG images, MP4 and WebM videos, GIFs, and JSON-based Lottie animations. Folders are not currently available inside the Media Library, so descriptive filenames are important.

Use names such as:

• Dashboard – Account Settings
• Product Page – Checkout Button
• Company Logo – White
• Tutorial Intro Animation
• Mobile App – Login Screen

Keep Recordings Short and Focused

Synthesia recommends keeping screen recordings under approximately 30 minutes for smoother recording and processing. Longer tutorials should usually be divided into separate lessons or chapters.

A focused recording is easier to update when the software changes. Instead of producing one 30-minute walkthrough, create several shorter videos covering individual tasks.

Review the Complete Demonstration

Before continuing, preview every recording and confirm that:

• No passwords or personal information are visible.
• Cursor movements are easy to follow.
• The correct window or tab was captured.
• The transcript matches the narration.
• Filler-word removal did not damage sentences.
• The recording begins and ends naturally.
• Text does not cover important controls.
• The avatar does not block the interface.
• The playback speed remains comfortable.
• Each scene demonstrates one clear step.

A screen recording should show the action while the narration explains why it matters. Combining clear cursor movement, short scenes, accurate text, and focused voiceover makes the tutorial easier to understand.

Add Captions, Music, Branding, and Transitions

Captions, background music, brand elements, and scene transitions help make a Synthesia video easier to understand and more visually consistent. Add these features after the main script, scenes, avatar, and visuals are mostly complete so later structural edits do not require repeated adjustments.

Understand the Caption Options

Synthesia currently provides three main caption types:

Closed captions can be turned on or off by viewers in the Synthesia video player. They match the language and wording of the script and cannot be customized with different fonts, sizes, or colors.

Dynamic captions appear directly on the canvas as editable visual elements. They can use custom fonts, colors, shadows, active-word highlighting, and word-by-word animations.

Burned-in captions are permanently added to the generated video. Viewers cannot turn them off, and their appearance cannot be customized.

Add Dynamic Captions

Dynamic captions are useful for social videos, branded explainers, and other projects where the subtitles should be part of the visual design.

1. Select the scene you want to edit.
2. Open Captions from the top toolbar.
3. Choose a caption preset.
4. Select the caption box on the canvas.
5. Adjust the font, size, color, shadow, and animation.
6. Move or resize the caption box.
7. Use script triggers when different caption sections should appear at specific moments.
8. Preview the scene.

The default dynamic-caption preset can display the scene’s complete script, while triggers provide more control over when individual portions appear. Dynamic captions can also use fonts and colors from an applied Brand Kit.

Keep Captions Readable

Place captions where they do not cover the avatar’s face, product images, interface buttons, charts, or other important visual information.

Use a font that remains readable on mobile devices and choose colors with strong contrast against the background. A shadow or background shape can help when footage changes between bright and dark scenes.

Avoid showing a full paragraph at once. Divide long narration into shorter scenes or caption sections so viewers have enough time to read each line.

Correct Caption Mistakes

Closed captions and burned-in captions follow the written script. To correct a spelling mistake, name, number, or punctuation error, edit the script itself.

When a word must appear correctly in captions but needs to be spoken differently, keep the official spelling in the script and use Synthesia’s pronunciation tool to control how the narrator says it.

For example, the captions can display a company’s proper name while the pronunciation control provides a phonetic version for the AI voice.

Add Burned-In Captions

Burned-in captions are added during final video generation.

1. Click Generate.
2. Turn on Burned-in Captions.
3. Review the remaining generation settings.
4. Generate the video.
5. Check the finished version.

These captions become a permanent part of the video and cannot be disabled by viewers. Because their styling is fixed, use dynamic captions instead when precise branding or positioning is required.

Download Subtitle Files

After the video has been generated:

1. Open the video’s three-dot menu.
2. Select Download Subtitles.
3. Choose .srt or .vtt.
4. Save the file.
5. Review it before uploading it elsewhere.

These files can be used when publishing the video on platforms that support separate subtitle uploads.

Captions always match the language of the video’s script. Translating the script creates captions in the translated language rather than keeping captions in a separate language.

Add Background Music

Music can establish the tone and make transitions between scenes feel smoother.

To add it:

1. Select a scene.
2. Turn on Music from the right-hand toolbar.
3. Choose a source from the Add Music panel.
4. Browse or search for a track.
5. Use the available filters.
6. Preview the track.
7. Click Add.
8. Adjust the volume.
9. Select Change All when the same music should apply throughout the video.

Synthesia allows users to choose music from their workspace library, reuse tracks already added to the project, or browse the integrated Soundstripe library. Tracks can be filtered by characteristics such as genre, mood, duration, and whether they are instrumental.

Balance the Music Volume

Synthesia applies new background music at 10% volume by default and recommends keeping it at approximately 15% or lower so it does not overpower the avatar’s voice.

Treat that recommendation as a starting point rather than a rule. Some quiet tracks may need a slightly higher setting, while songs with loud drums, vocals, or strong bass may need to be lowered further.

Listen to several scenes containing narration. The speaker should remain easy to understand without requiring the viewer to concentrate.

Use Different Music in Separate Scenes

Synthesia allows different tracks to be applied to individual scenes. Select the scene and choose another song from the Music panel.

Different music can help separate an introduction, main lesson, demonstration, and conclusion. However, changing tracks too frequently can make a short video feel disconnected.

For most beginner projects, one consistent background track is enough. Use another track only when the tone or purpose of the section clearly changes.

Upload Your Own Music

To upload a custom track:

1. Open the Add Music panel.
2. Select Upload Music.
3. Choose the audio file.
4. Wait for the upload to finish.
5. Add it to the selected scene.
6. Adjust its volume.
7. Apply it to other scenes when needed.

Synthesia currently supports uploaded music in MP3, WAV, OGG, AAC, and FLAC formats. Uploaded audio can range from three seconds to five minutes per scene.

Only upload music you own or have permission to use. Saving a track to the workspace does not create a license for copyrighted audio.

Review Music Across Scene Changes

Play the full video and listen for sudden changes in volume or mood. A track that works during the introduction may feel too energetic during detailed instructions or serious information.

Check whether:

• The music begins naturally.
• The narration remains clear.
• Track changes happen at logical points.
• Music does not end abruptly.
• The final scene has enough time to finish cleanly.

Use silence when music would reduce clarity. A software demonstration, safety warning, or important statement may be stronger without background audio.

Add Branding Manually

Users who do not have access to Brand Kits can still brand a video manually by adding a logo, choosing consistent colors, and using the same fonts and layouts throughout the project.

Add a transparent PNG or SVG logo through the Media menu, place it in a consistent corner, and reuse the same size and position across scenes. Check that it does not overlap captions, avatars, or important content.

Choose a small set of colors and text styles before designing the complete project. This prevents scenes from looking as though they belong to different videos.

Create a Brand Kit

Synthesia’s current Brand Kit feature is limited to Enterprise plans. It allows organizations to store approved logos, text styles, avatars, colors, themes, and supported avatar outfits.

To create one manually:

1. Open Assets.
2. Select Brand Kits.
3. Click Create Brand Kit.
4. Enter a name.
5. Add a primary logo for lighter backgrounds.
6. Add a negative logo for darker backgrounds.
7. Select the approved text styles, avatars, and colors.
8. Create or customize the color themes.
9. Click Save.

A Brand Kit can currently contain up to two text styles, six avatars, and 120 colors. Each theme contains four colors: two light and two dark.

Generate a Brand Kit From a Website

Enterprise users can also enter a brand name or website address and allow Synthesia to generate an initial Brand Kit.

1. Open Brand Kits under Assets.
2. Select Create Brand Kit.
3. Choose the automated creation option.
4. Enter the website or brand.
5. Review the extracted assets.
6. Correct the colors, logos, and text styles.
7. Upload any custom fonts manually.
8. Save the finished kit.

Custom fonts are not automatically imported when a Brand Kit is generated from a website, so they must be uploaded and selected separately.

Always compare the generated kit with the company’s actual brand guidelines. A website may contain temporary campaign colors, secondary logos, or fonts that are not approved for general use.

Apply a Brand Kit

To apply an existing Brand Kit:

1. Open the project.
2. Select the video action menu beside the project title.
3. Choose the Brand Kit.
4. Review how the scenes change.
5. Select different approved themes for individual scenes when needed.
6. Correct any layouts or contrast problems.

Applying a Brand Kit updates the project’s color palette across its scenes and places the approved color choices first when editing visual elements.

The Brand Kit can also be applied to templates. Users without an Enterprise plan can still customize the colors, fonts, and logo manually inside a template.

Review the Branding

After applying branding, check every scene. An approved color may still provide poor contrast against a particular image, while the normal logo position may cover a presenter or screenshot.

Confirm that:

• The correct logo variation appears.
• The logo remains visible against the background.
• Fonts are readable.
• Colors match the official brand.
• Caption styling remains clear.
• Avatar clothing and environments feel appropriate.
• Opening and closing scenes use consistent visual treatment.

Brand consistency should make the video recognizable without overcrowding it with logos and repeated company information.

Add Scene Transitions

Transitions control how one scene changes into the next.

1. Select the scene where the transition should begin.
2. Turn on Scene Transition from the right-hand toolbar.
3. Choose a transition style.
4. Select Change All when the same transition should be used throughout the project.
5. Preview the complete video.

Synthesia’s transitions can be applied to one scene or copied across the full video. A Direct Cut means the next scene begins immediately without a transition effect.

Preview Transitions Correctly

Scene transitions do not appear in individual scene previews. Preview the complete video to see how they look between scenes.

Transitions are not available on the final scene or on scenes containing interactive elements.

Watch several scene changes in sequence. A transition may look acceptable once but become repetitive or distracting when used throughout a longer video.

Use Transitions Selectively

A direct cut is often the clearest choice for tutorials, fast explanations, and scenes that continue one idea. A smoother transition may work better when the video moves to a new chapter, speaker, location, or subject.

Avoid using several unrelated transition styles in one project. Choose one main style and use a second option only when an important section change needs stronger separation.

Complete a Final Review

Before moving on, preview the complete video and confirm that:

• Captions match the script.
• Dynamic captions are readable and correctly positioned.
• Music remains quieter than the narration.
• Uploaded music is properly licensed.
• Logos and colors remain consistent.
• Brand elements do not cover important visuals.
• Scene transitions feel smooth.
• Transition styles are not overused.
• The introduction and conclusion begin and end naturally.

Captions should improve understanding, music should support the narration, branding should create consistency, and transitions should connect scenes without drawing unnecessary attention to themselves.

Translate a Video Into Another Language

Synthesia provides two localization workflows. Video Translation creates translated versions of a video originally made in Synthesia, while AI Dubbing translates the spoken audio in an uploaded video or YouTube video. These tools serve different purposes and have different plan requirements.

Choose Between Translation and Dubbing

Use Video Translation when the original project was created inside Synthesia and you want to translate its script, narration, and eligible on-screen text. Synthesia’s one-click translation workflow and multilingual video player are currently limited to Enterprise plans.

Use AI Dubbing when you already have a finished video file or a YouTube video containing spoken audio. Dubbing is currently available on all plans and uses account credits.

The main difference is:

• Translation works with an editable Synthesia project.
• Dubbing works with the spoken audio of an existing video.

Choose the workflow before making major localization changes, since dubbing does not automatically translate titles, graphics, or other text already embedded in the footage.

Translate a Synthesia Project

To translate a video created in Synthesia:

1. Locate the original project.
2. Select Translate from the project page or video menu.
3. Open the language dropdown.
4. Choose one or more target languages.
5. Review the available translation settings.
6. Decide whether to translate only the script.
7. Decide whether Synthesia should generate each version automatically.
8. Select Translate.
9. Follow the progress indicator.
10. Open the new translations folder when processing finishes.

Synthesia creates a folder named after the original project and stores the source video and translated versions together.

Choose What Should Be Translated

The Only translate the script setting keeps the project’s on-screen text in its original language while translating the spoken script and narration. Leave this option disabled when visible titles and supported text elements should also be localized.

Translate only the script when the on-screen wording includes:

• A product name that should remain unchanged
• An internationally recognized slogan
• A screenshot that cannot be edited
• Legal wording that will be replaced manually
• Text intended for several language audiences

Translate the visual text as well when the video contains instructions, headings, labels, calls to action, or other information viewers need to understand.

Decide Whether to Generate Automatically

Turn on Automatically generate videos when the translations should be rendered immediately after they are created.

Leave it off when a fluent speaker or localization reviewer needs to check the translated scripts first. Without automatic generation, each translated version must be generated manually before it can be published or shared.

Manual review is usually safer for:

• Legal or compliance videos
• Technical instructions
• Medical or safety information
• Product terminology
• Customer-facing marketing
• Content using humor or regional expressions

Automatic generation may be more practical for simple internal updates or low-risk content that has already been standardized.

Review the Translated Script

Open each language version and read the complete script before generating it. Automated translation can preserve the general meaning while still producing unnatural wording, incorrect terminology, or a tone that does not fit the audience.

Ask a fluent speaker to check:

• Meaning and factual accuracy
• Grammar and natural phrasing
• Formality and tone
• Product and company terminology
• Regional vocabulary
• Numbers, dates, and measurements
• Pronunciation of names and abbreviations

Synthesia recommends previewing non-English voices and having pronunciation-sensitive scripts checked by a native speaker.

Choose the Correct Regional Voice

A language may include several regional variants. For example, Spanish, English, French, and Portuguese can have different voices or accents depending on the intended country or audience.

Open the speaker controls in the translated project and confirm that the selected voice uses the appropriate language and regional variation. Synthesia currently provides stock voices across more than 140 languages and accents, but voice and dialect availability varies.

A Spanish translation intended for Spain may need a different voice than one intended for Mexico or the United States. The written translation may also require different vocabulary even when the language is technically the same.

Review Text Length and Scene Timing

Translated sentences may be shorter or longer than the original wording. This can change the narration length, scene timing, animations, and amount of text displayed on screen.

After translating:

1. Preview every scene.
2. Check whether the narrator finishes too early or late.
3. Shorten text that no longer fits the layout.
4. Reposition titles and captions.
5. Adjust animation triggers.
6. Divide overcrowded scenes when necessary.
7. Confirm that media remains visible for the correct amount of time.

Do not force a long translation into a small text box by reducing the font excessively. Adjust the layout or divide the information instead.

Translate With an XLIFF File

Enterprise users can export a project’s script as an XLIFF file, translate it through a professional Translation Management System, and import the translated file back into Synthesia. XLIFF preserves the translation text, structure, contextual information, and inline markers used by localization tools.

This workflow is useful when an organization already works with:

• Professional translators
• Translation agencies
• Approved terminology databases
• Localization software
• Formal review and approval processes

After importing the translated XLIFF file, review the project inside Synthesia to confirm that the text still fits its scenes and layouts.

Update Translations After Editing the Original

Changes made to the original video do not automatically appear in its existing translations. After editing and regenerating the source project, the translated versions must be refreshed.

To update them:

1. Edit the original project.
2. Generate the revised original video.
3. Select the language icon beside its title.
4. Choose Add Translation.
5. Select the languages that need updating.
6. Translate them again.
7. Review each updated version.
8. Regenerate and republish the translations.

Synthesia keeps the previous translated versions as drafts until the revised ones are generated and published.

Complete the original project as thoroughly as possible before translating it. Repeated source changes create additional review work across every language.

Share a Multilingual Video

Enterprise users can publish several translated versions through one multilingual video player.

1. Open the original generated video.
2. Select Share.
3. Turn on Include Translations.
4. Add the translated versions.
5. Create the sharing link or embed code.
6. Test the language selector.

Viewers can then choose their preferred language from the selector in the video player.

This approach is useful for an international website, company intranet, learning platform, or knowledge base because one player can contain several language options.

Share One Translation Separately

A translated version can also be shared independently when a particular audience should receive only one language.

To create a separate version:

1. Find the translation in the project folder.
2. Open More Actions.
3. Select Duplicate.
4. Open the duplicate.
5. Generate it as an independent video.
6. Select Share.
7. Create a dedicated link or embed code.

The duplicate is no longer connected to the original project. Future source edits and translations will not update it automatically.

Use a separate version when distributing content to one regional office, customer group, course, website, or language-specific campaign.

Dub an Existing Video

AI Dubbing can translate the spoken audio from a video uploaded to Synthesia or provided through a YouTube link.

1. Select Dubbing from the main sidebar.
2. Upload the video or paste a YouTube link.
3. Choose one or more target languages.
4. Open Advanced Options.
5. Confirm or change the original language.
6. Enter a clear project name.
7. Choose whether to enable lip synchronization.
8. Select the duration behavior.
9. Choose Dub Video.
10. Monitor the translation progress.

Synthesia creates a folder under My Videos containing the original and dubbed versions. Videos dubbed from a YouTube link cannot currently be published by users on Basic, Starter, or Creator plans.

Review the Transcript Before Dubbing

Select Review Transcript when you want to correct the original transcription before spending credits on translated versions. Reviewing the source transcript does not consume dubbing credits.

Check the transcript for:

• Speaker names
• Product names
• Technical terminology
• Numbers and prices
• Abbreviations
• Background speech
• Words affected by poor audio quality

Translation quality depends on the source transcript. An incorrectly transcribed company name will probably remain incorrect in every dubbed language unless it is corrected first.

Enable Lip Synchronization

Turn on Lip Sync when the original video visibly shows people speaking and their mouth movements should match the translated narration.

Lip synchronization may make a presenter-led video feel more natural, but it increases credit consumption and requires additional processing. When the speaker’s mouth is not visible, such as during screen recordings, presentations, product footage, or voiceover-only videos, lip synchronization may be unnecessary.

Review the final result for unusual mouth movement, facial distortion, or sections where the translated audio does not align naturally with the speaker.

Choose the Video-Duration Setting

Synthesia currently provides two timing options for dubbing.

Adaptive adjusts the video playback to better accommodate the translated narration. Synthesia recommends this option for instructional material.

Original preserves the original video timing and adjusts the translated voiceover instead. This may be better for persuasive or visually timed content where the footage must retain its original pacing.

Use Adaptive for tutorials, training videos, and demonstrations where clarity matters most. Use Original when music, edits, actions, or visual transitions depend closely on the existing timing.

Understand What Dubbing Does Not Translate

AI Dubbing translates spoken audio. It does not automatically translate text already visible inside the original footage, including:

• Burned-in captions
• Titles
• Presentation slides
• Product labels
• Buttons and interface text
• Charts or graphics

Update this text in the source video before dubbing or create separate localized visual versions.

Do not leave important instructions in one language while translating only the narration. Viewers may hear the correct instruction but still see buttons or labels they cannot understand.

Review Every Language Version

Before publishing, have each translated video checked by someone who understands the language and intended audience.

Confirm that:

• The translation preserves the original meaning.
• The regional vocabulary is appropriate.
• The voice sounds natural.
• Names and terminology are pronounced correctly.
• Captions and on-screen text match the narration.
• Text still fits inside the layouts.
• Animations remain synchronized.
• Lip movements appear acceptable.
• Calls to action and links lead to the correct regional destination.

Translation should adapt the video for its audience rather than simply replace individual words. A careful review of language, timing, design, and cultural context makes the localized version feel like an intentional video instead of an automatic copy.

Preview, Generate, Share, and Download the Finished Video

Preview the complete project before generating it. Synthesia’s preview tools allow users to check the script, visual timing, layouts, and animations without consuming credits. The avatar’s complete movement and lip synchronization do not appear until the video is generated.

Preview One Scene

To review an individual scene:

1. Select the scene from the left-hand panel.
2. Click the Preview Scene play button near the script box.
3. Watch the script, media, and animations from the beginning.
4. Move along the seeker bar when you want to preview from a specific point.
5. Correct any problems and preview the scene again.

For more precise timeline movement, hold Ctrl on Windows or Command on a Mac while scrolling.

Scene previews are useful for checking small changes without replaying the entire video.

Preview the Complete Video

Click the play button near Generate in the upper-right corner of the editor to preview the full project.

During the preview, check that:

• Scenes appear in the correct order.
• The narration sounds natural.
• Visuals support the script.
• Text remains readable.
• Animations appear at the correct time.
• Screen recordings are easy to follow.
• Music does not overpower the voice.
• Transitions feel consistent.
• The conclusion does not end abruptly.

The preview does not show the avatar’s final animation, so pronunciation, script rhythm, and avatar placement should be reviewed separately before generation.

Review the Script One Final Time

Correcting the script before generation reduces the need to spend additional credits on another version.

Check names, numbers, dates, product details, technical terms, and calls to action. Make sure no placeholder wording remains and confirm that every scene uses the correct speaker, language, and voice.

For self-service plans, each generated second currently uses two credits. A one-minute video therefore uses 120 credits. Editing and regenerating a project can use additional credits when the revised version contains newly generated duration.

Check the Credit Balance

Before generating:

1. Open Account Settings.
2. Select Billing.
3. Find the Usage section.
4. Review the remaining credit balance.
5. Compare it with the estimated cost of the video.

Synthesia also displays usage warnings in the main menu and during generation when a project would exceed the available balance.

When an account does not have enough credits, the user must wait for the next renewal or upgrade the plan before generating the video.

Generate the Video

When the project is ready:

1. Click Generate in the upper-right corner.
2. Enter or review the video title.
3. Add a description when needed.
4. Turn on burned-in captions when they should remain permanently visible.
5. Review or edit the chapter settings.
6. Click Generate again.
7. Follow the processing status until the video is complete.

Synthesia may display stages such as moderation, queueing, preparation, rendering, assembly, and finalization while processing the project.

Users can currently have up to ten videos generating at once. Additional projects remain queued until one of the active generations finishes.

Understand Content Moderation

A project may enter a moderation stage before rendering. Synthesia checks generated content against its platform rules, including restrictions related to impersonation, deceptive material, unsafe content, and unauthorized uses of avatars or voices.

Make sure all personal avatars, cloned voices, photographs, recordings, and branded materials are used with the required permission. When generation fails, review the displayed error message before changing the project.

Review the Generated Video

Once generation finishes, play the full rendered version. This is the first time the avatar’s complete facial animation, body movement, gestures, and lip synchronization can be reviewed accurately.

Check for:

• Mispronounced words
• Unnatural pauses
• Incorrect avatar movements
• Lip-synchronization problems
• Text appearing too early or late
• Cropped images
• Abrupt audio changes
• Weak scene transitions
• Caption errors
• Awkward final timing

Do not publish the video based only on the editor preview.

Edit a Generated Video

Select Open in Editor when the generated result needs corrections. Make the changes and generate the project again.

Each generation creates a separate version of the project. Synthesia’s version history records the generated state, including its script, visual elements, animations, and other project settings.

This allows users to review earlier versions and maintain an existing published link while updating the video.

Name Each Version Clearly

Use descriptive project titles and version notes when several revisions exist.

For example:

• Employee Onboarding – Draft 1
• Employee Onboarding – Reviewed
• Employee Onboarding – Final
• Employee Onboarding – Spanish

Clear names reduce the chance of sharing or downloading an outdated version.

Create a Shareable Link

To share the video through Synthesia:

1. Open the generated video.
2. Click Share.
3. Select Create Link.
4. Copy the public link.
5. Test it in another browser before distributing it.

The link opens a Synthesia-hosted video page. Deleting the link stops public access through that address.

Customize the Shared Video Page

Depending on the account and video, sharing settings may allow users to:

• Let viewers duplicate the video into their own Synthesia workspace.
• Add a clickable call-to-action with a label and destination.
• Copy an animated GIF thumbnail.
• Create an embed code.
• Protect access with a password.
• Require organizational single sign-on.

The call-to-action option works on the Synthesia webpage version of the shared video.

Only enable duplication when viewers should be allowed to create their own editable copy.

Embed the Video on a Website

Select Share, create a link, and copy the embed code. Place that code into the website, learning platform, knowledge base, or supported content-management system.

Videos embedded from Synthesia can display later generated updates without requiring users to upload a new video file to every webpage.

Test the embedded video on desktop and mobile screens. Confirm that it loads correctly, fits its container, and does not interfere with other page elements.

Share a GIF Thumbnail

Synthesia can generate an animated GIF thumbnail that links viewers to the hosted video page. This can be useful in emails, internal messages, and sales outreach because recipients can see a short visual preview before opening the full video.

The GIF is a preview rather than the complete video. Make sure its connected link leads to the correct published version.

Export for a Learning Platform

Enterprise users can export generated videos as SCORM 1.2 or SCORM 2004 packages for compatible learning-management systems.

1. Open the generated video.
2. Select Share.
3. Create the share link.
4. Find Download SCORM.
5. Choose the SCORM version.
6. Set the percentage viewers must watch for completion.
7. Download the ZIP package.
8. Upload it to the learning platform.

SCORM packages can report completion information based on the selected viewing requirement.

Download the MP4

To save the video to a device:

1. Make sure the project has been generated.
2. Open the three-dot menu in the upper-right corner.
3. Select Download.
4. Choose MP4.
5. Save the file with a recognizable name.
6. Play the downloaded version before publishing it.

Synthesia currently downloads MP4 videos in Full HD 1080p at 1920 × 1080 resolution.

Video downloading is not included with the Basic plan. Starter, Creator, and eligible Enterprise accounts provide download access.

Download Audio and Subtitle Files

The download menu can also provide:

• WAV for the generated audio
• SRT for subtitles
• VTT for web captions
• XLIFF for supported translation workflows

These files are available after the video has been generated.

Review subtitle files before uploading them elsewhere. They follow the project’s script, so any remaining spelling or punctuation mistakes may also appear in the exported captions.

Understand Interactive-Video Restrictions

Videos containing buttons, branching, or other interactive elements cannot be downloaded as standard MP4 files. Interactive functionality depends on viewer input and cannot be preserved inside a normal video file.

Interactive videos can instead be:

• Shared through a Synthesia link
• Embedded on a website or learning platform
• Exported as SCORM on eligible Enterprise accounts

To create a downloadable alternative, duplicate the project, remove its interactive elements, and generate the noninteractive version.

Review the Downloaded File

Play the downloaded MP4 from beginning to end and confirm that:

• The picture is clear.
• The audio remains synchronized.
• Captions stay inside the frame.
• The avatar looks natural.
• Music volume is balanced.
• No interface elements are cropped.
• The correct version was downloaded.
• The final scene remains visible long enough.
• The file opens on the intended platform.

Also test shared links, embeds, calls to action, password protection, and SCORM behavior separately. A local MP4 review cannot confirm whether those online features work correctly.

Once the rendered video, sharing settings, and downloaded files have been reviewed, the project is ready to distribute to its intended viewers.

Manage Projects, Credits, and Plan Limits

Synthesia saves projects under My Videos or the selected workspace. Organizing these projects and monitoring credit usage can prevent lost work, accidental edits, and interruptions during video generation.

Rename Projects Clearly

Rename each project as soon as it is created. Descriptive titles make drafts and generated versions easier to identify.

A useful naming structure might be:

• Customer Onboarding – Main Version
• Customer Onboarding – Spanish
• Product Tutorial – Vertical Short
• Compliance Training – July 2026
• Software Demo – Draft 2

Include the topic, language, format, audience, or version when those details help distinguish similar projects.

Create Project Folders

Folders can organize projects by client, department, course, campaign, language, or publication status.

To create one:

1. Open My Videos.
2. Select Create Folder.
3. Enter the folder title.
4. Open the folder to create subfolders when needed.
5. Drag projects into the appropriate location.

Folders can be renamed from their three-dot menus. Synthesia also allows users to select several videos and move them together through the Move To action.

A simple folder structure could include Drafts, Under Review, Published, and Archived.

Move a Video

To move a project:

1. Find the video.
2. Open its three-dot menu.
3. Select Move To.
4. Choose the destination folder.
5. Create a new folder when necessary.
6. Select Move.

On eligible Enterprise workspaces, users may also be able to move videos between their private My Videos area and a shared workspace. Projects stored in a shared workspace are visible according to workspace access, while projects under My Videos remain private to their owner.

Duplicate a Project

Create a duplicate before making major changes to an approved project.

1. Open My Videos.
2. Find the project.
3. Open its three-dot menu.
4. Select Duplicate.
5. Rename the copied version.
6. Edit the copy while keeping the original unchanged.

Duplicating does not consume generation credits. Credits are used only when the duplicate is generated or when another credit-consuming feature is applied. Confirm that the cloud icon shows the latest edits are saved before creating the copy.

Duplicates are useful when creating:

• Another language version
• A shorter social-media edit
• A version for a different audience
• A new presenter or voice variation
• An updated course or product tutorial

Do not edit the original approved project when the new version may require substantial experimentation.

Confirm That Edits Are Saved

Synthesia saves project changes automatically while the editor remains connected to the platform.

The cloud icon shows the current save status:

• A checkmark means the edits are saved.
• An upward arrow means saving is still in progress.
• A red X means the project is not currently saving.

When the red X appears, stop editing and wait for the connection to return. Changes made while saving is unavailable may be lost.

Before closing the browser, duplicating a project, or switching workspaces, confirm that the cloud displays a checkmark.

Use Version History

Synthesia creates a new generated version each time Generate is selected. It also saves an automatic project snapshot approximately every 30 minutes.

To open the history:

1. Open the project in the editor.
2. Select the menu beside the project title.
3. Choose Show Version History.
4. Review the versions and automatic snapshots.
5. Open the three-dot menu beside an earlier version.
6. Choose Restore or Duplicate.

Restore replaces the current working version with the selected version. Duplicate creates a separate copy, allowing users to preserve both versions.

Duplicating is usually safer when you are unsure whether the earlier version should replace the current one.

Update a Published Video

When a newer generated version should replace an older published version, open version history, choose the newer version, and select Reshare, followed by Replace.

The existing share page, embed, and eligible SCORM distribution can update to the new version without changing the original sharing URL. Viewers using the existing link will then see the replacement version.

This is useful for updating software tutorials, company policies, course lessons, or product information without replacing links across several websites and documents.

Delete Projects Carefully

Deleting a video moves it into Synthesia’s trash rather than removing it immediately. The video can be restored for up to 30 days. After that period—or after manual permanent deletion—it is removed and cannot be recovered.

Deleting a video does not return the credits used to generate it. It can also separate a translated version from its original translation set.

Before deleting an important project, save the finished MP4, subtitles, script, original media, and any other files needed outside Synthesia.

Understand Synthesia Credits

Basic, Starter, and Creator accounts use a shared credit balance for generation-based features. Regular video generation currently consumes two credits for every generated second, meaning a one-minute video uses 120 credits.

Credits may also be used for features such as AI-generated visuals, AI Dubbing, and other usage-based tools. The amount depends on the feature and selected model.

Before using an advanced feature, check the displayed estimate rather than assuming it costs the same as normal video generation.

Review the Current Plan Allowances

Synthesia currently lists the following standard self-service allowances:

• Basic: 1,200 credits per month
• Starter monthly: 1,200 credits per month
• Starter annual: 14,500 credits provided for the annual term
• Creator monthly: 3,600 credits per month
• Creator annual: 44,000 credits provided for the annual term

At the standard video-generation rate, 1,200 credits can cover up to approximately ten minutes of regular generated video, while 3,600 credits can cover up to approximately thirty minutes. Using credits for AI assets, dubbing, or other features reduces how much remains for normal video generation.

Because plan features and allowances can change, verify the current Pricing and Billing pages before upgrading.

Check the Credit Balance

To view usage:

1. Open Account Settings.
2. Select Billing.
3. Find the Usage section.
4. Review the remaining credits and renewal details.

Synthesia also displays alerts in the main menu and during generation when usage is approaching or exceeding the available balance. A project cannot begin generating when its estimated cost is higher than the remaining credit balance.

Check the balance before producing a long project, translating several versions, using AI-generated clips, or adding lip-synchronized dubbing.

Understand Monthly and Annual Credits

Monthly plans refresh their credit balance at the beginning of each billing period. Annual self-service plans currently release the annual credit allocation at the start of the subscription instead of dividing it into monthly portions.

Unused standard credits do not roll over into the next billing period. When the plan renews, the available balance resets according to that plan’s allowance.

Annual subscribers should still budget usage across the year. Receiving all credits at once makes them flexible, but it also makes it possible to consume a large portion early in the subscription.

Understand Credit Use After Editing

Regenerating an edited video does not always charge the entire original video again. Synthesia measures the newly generated material created by supported edits.

Adding new words increases the generated duration and consumes credits for that additional time. Replacing existing words is treated as removing the old words and generating the replacement. Removing words, changing punctuation, or rearranging existing wording generally does not add generation usage by itself.

Other changes, such as switching an avatar or voice, may require affected content to be regenerated. Review the displayed credit estimate before confirming the new generation.

Avoid Unnecessary Regeneration

Preview scenes and the complete project before clicking Generate. Previewing does not consume standard video-generation credits, while producing another rendered version may use additional credits for changed content.

Before generating, confirm that:

• The script is final.
• Names and terminology are pronounced correctly.
• The right avatar and voice are selected.
• Scenes are in the correct order.
• Captions and text contain no mistakes.
• Visuals are properly positioned.
• Music and transitions are correct.
• The desired language version is open.

A careful review is especially important when several translated versions must be generated.

Understand AI-Asset Costs

AI-generated images and video clips use the same shared credit balance on self-service plans. Costs vary by model.

For example, Synthesia’s current documentation lists different credit rates for supported image and video models, with advanced video-generation models costing considerably more than a standard image. The platform displays the expected amount before generation.

Create a simple test before generating several similar assets. Once the prompt, style, and model produce acceptable results, continue with the remaining scenes.

Understand AI Dubbing Costs

AI Dubbing consumes credits separately from ordinary video generation. Enabling lip synchronization increases the cost because Synthesia must also modify visible mouth movement. Its current documentation lists one minute of lip-synchronized dubbing at 240 credits.

Do not enable lip synchronization when the speaker’s mouth is not visible. Screen recordings, slide presentations, product footage, and voiceover-only content may not benefit from the additional processing.

What Happens When Credits Run Out?

When the available balance is insufficient, Synthesia prevents the video or credit-based feature from generating.

Users on Basic, Starter, or Creator plans must either wait for the next billing renewal or upgrade to a plan with a larger allowance. When the account appears to have enough credits but still displays an error, Synthesia recommends contacting support.

Shortening the video can reduce the required amount, but removing important information only to fit the balance may weaken the final project.

Understand Plan Upgrades

A self-service plan can be upgraded from Settings, Billing, and Upgrade. Synthesia applies the upgrade immediately and calculates a prorated charge based on the current billing cycle.

Unused credits from the previous plan do not transfer into the upgraded plan’s balance. Instead, the account resets to the new plan’s full allowance, and the renewal date may change according to the prorated transaction.

Review the unused balance before upgrading so valuable credits are not abandoned unexpectedly.

Understand Workspace Limits

Self-service workspaces are designed around one primary user and may include guest access. Shared team workspaces, broader collaboration controls, folder permissions, and organization-level management are associated with Enterprise plans.

Enterprise administrators can allocate credits to different workspaces, monitor usage, and receive alerts when consumption reaches specified levels. When workspace limits are enabled, administrators can decide which teams receive particular allowances and which credit-consuming features are available.

Smaller users usually need only the default workspace. Multiple workspaces become more useful when departments require separate permissions, budgets, assets, or content libraries.

Share Folders Carefully

Enterprise folder sharing allows workspace members or invited guests to receive viewing, commenting, editing, or broader access according to the selected permission. External guests see only folders explicitly shared with them.

Before sharing a folder, confirm that it does not contain unrelated drafts, internal scripts, confidential media, or personal avatars that the recipient should not access.

Keep External Backups

After finishing an important video, save:

1. The generated MP4
2. The script
3. SRT or VTT subtitles
4. Uploaded images and video clips
5. Music and narration files
6. Translation files when applicable
7. Any approvals or licensing documentation

Synthesia’s version history and trash provide useful recovery options, but external backups protect completed work from account changes, accidental deletion, expired access, or subscription problems.

Clear project names, organized folders, saved versions, and regular credit checks make Synthesia easier to manage. They also reduce the likelihood of generating the wrong project, losing an approved version, or unexpectedly reaching the plan limit.

Tips for Using Synthesia Effectively

Synthesia can automate narration, avatar animation, layouts, translation, and other production tasks, but its first result should still be treated as a draft. Careful scripting and review usually have a greater effect on quality than adding more visual effects.

Begin With One Clear Objective

Decide what viewers should understand or do after watching the video.

A focused objective might be:

“Show new employees how to submit an IT support request correctly.”

This is more useful than a broad objective such as:

“Explain the IT department.”

Use the objective to decide which scenes, examples, and visuals belong in the video. Remove information that does not help viewers reach the intended outcome.

Define the Audience

Consider what viewers already know, what terminology they understand, and why they are watching.

A tutorial for first-time users should explain buttons and basic navigation. A video for experienced employees can move more quickly and focus on changes, exceptions, or advanced procedures.

The avatar, voice, language, tone, examples, and pacing should all fit the intended audience.

Start With a Short Practice Video

Create a 30- to 60-second project before building a long training course or presentation. Use the test to learn how scenes, avatars, voices, pronunciation, layouts, previews, and generation work.

A short project also lets users evaluate the final avatar animation without spending a large portion of their credits. Standard self-service generation currently uses two credits for each generated second.

Treat Assistant as a Starting Point

Synthesia’s Assistant can turn a prompt and supporting files into a first draft, revise scripts, and update visual elements inside the editor. However, its output should still be checked for accuracy, unnecessary wording, unsupported claims, and generic examples.

Provide Assistant with specific instructions about:

• The audience
• The video’s purpose
• Required information
• Preferred length
• Desired tone
• Information it should avoid adding

The more clearly the task is explained, the less rewriting the generated draft may require.

Write for Listening

Use short sentences and natural spoken language. A script that reads well as a report may still sound formal or difficult when spoken aloud.

Instead of:

“The following section will provide an explanation of the process that employees are required to complete when requesting technical assistance.”

Use:

“Next, let’s see how to request technical support.”

Synthesia scripts are organized by scene, and longer scripts should be divided across several scenes. Scene breaks also create natural pauses in the narration.

Keep One Main Idea in Each Scene

A scene should normally explain one instruction, example, comparison, or transition.

Divide a scene when:

• The topic changes
• Another step begins
• A different visual is needed
• The presenter moves from explanation to demonstration
• The on-screen text becomes crowded

Avoid splitting every short sentence into its own scene. Scenes should change because the message or visual focus changes, not simply because another sentence begins.

Reach the Main Point Quickly

Avoid long introductions that explain what the video will eventually discuss. Tell viewers what they will learn, why it matters, and then begin.

For example:

“In this video, you’ll learn how to reset your password without contacting support. Let’s begin from the login page.”

This is usually stronger than spending several scenes discussing the importance of password security before showing the actual process.

Use Avatars Selectively

An avatar can create consistency and provide a recognizable presenter, but it does not need to remain visible throughout the entire video.

Use the presenter during introductions, explanations, transitions, and conclusions. Hide it during detailed screen recordings, diagrams, charts, or product demonstrations when viewers need the entire canvas.

The editor does not display complete avatar animation or lip movement during previews. Those elements appear after generation, so avatar placement and narration must be checked before rendering and the complete performance reviewed afterward.

Write Naturally for Expressive Avatars

Expressive avatars respond to wording, punctuation, pauses, and emotional cues in the script. Questions, encouraging language, exclamation points, and short pauses can influence the delivery.

For example:

“You completed the first step. Great job! Now, let’s connect your account.”

This may produce a more natural result than:

“The first step has been completed and the next step is account connection.”

Do not exaggerate punctuation throughout the complete script. Strong emotional cues should be reserved for moments that genuinely require enthusiasm, concern, curiosity, or emphasis.

Preview the Actual Voice

Listen to the chosen voice with the real script rather than relying only on its default sample. A voice may sound appropriate during a short demonstration but handle the project’s terminology, pacing, or emotional tone differently.

Synthesia’s Voices page allows users to test custom text with stock voices before applying one to a project.

Check sentences containing:

• Names
• Acronyms
• Product terminology
• Numbers
• Questions
• Foreign words
• Important warnings

Correct Pronunciation Before Timing Scenes

Use Synthesia’s pronunciation controls for company names, technical terms, acronyms, and other difficult wording. Users can enter a phonetic spelling or record the desired pronunciation and preview it before saving.

Correct these problems before finalizing animation triggers and scene duration. Changing the wording, pauses, or pronunciation can affect the narration’s timing.

Store frequently used terms in the pronunciation dictionary when they need to remain consistent across several videos.

Use Pauses Carefully

Add brief pauses after questions, statistics, major instructions, and transitions. Pauses can improve pacing and give viewers time to absorb information.

Do not place a noticeable pause after every sentence. Too many pauses can make the narrator sound disconnected and unnecessarily extend the video.

Scene breaks can also provide natural spacing when the topic or visual changes.

Show What the Narration Describes

The visual should support the exact information being spoken.

When the narration says, “Select Billing from the left-hand menu,” show the actual menu rather than generic footage of someone using a computer.

Use:

• Screen recordings for software instructions
• Product images for product explanations
• Charts for numerical comparisons
• Diagrams for processes
• Presenter scenes for introductions and summaries
• Full-screen text only for brief statements

Relevant visuals improve understanding more than decorative footage.

Keep On-Screen Text Brief

Do not place the full narration on the canvas unless viewers genuinely need to read every word.

Use on-screen text for:

• Titles
• Important terms
• Short steps
• Statistics
• Names and roles
• Warnings
• Calls to action

When a paragraph does not fit comfortably, shorten it or divide the content into additional scenes instead of reducing the font size excessively.

Design for Mobile Viewing

Even when the video is created on a desktop, many viewers may watch it on a phone or inside a small webpage player.

Check that:

• Text remains large enough to read
• Captions use strong contrast
• Buttons and screenshots are visible
• The avatar does not cover important content
• Logos are recognizable without being oversized
• Several small elements are not competing for attention

Preview the finished video at a reduced window size before publishing it.

Limit Decorative Animations

Use animations to reveal information at the moment it is discussed. For example, show each step as the narrator reaches it instead of displaying the full list immediately.

Avoid animating every title, icon, image, and shape simply because the option is available. Excessive movement can distract from the message and make educational content harder to follow.

Complete the script first, then use the Timeline to synchronize only the elements that benefit from precise timing.

Use Consistent Layouts

Choose a small group of layouts and repeat them throughout the project.

A simple video may use:

1. A title layout
2. A presenter-and-text layout
3. A full-screen media layout
4. A list or comparison layout
5. A closing layout

Reusing layouts gives the video structure while still allowing enough variation. Templates and duplicated scenes can reduce repetitive design work.

Keep Branding Consistent

Use the same logo position, colors, typography, caption design, and music direction across related videos.

Branding should make the project recognizable without covering the content. Check logo visibility over both light and dark backgrounds, and make sure brand colors still provide enough contrast for readable text.

Users without Enterprise Brand Kits can reproduce the same visual identity manually by reusing layouts, assets, and text styles.

Finish the Script Before Detailed Timing

Avoid spending time precisely timing animations and media while the narration is still changing.

A practical order is:

1. Finalize the script.
2. Confirm the avatar and voice.
3. Correct pronunciation.
4. Organize the scenes.
5. Add the main visuals.
6. Add captions and branding.
7. Adjust detailed animation timing.
8. Preview and generate.

This reduces repeated editing when a script change affects the scene duration.

Preview Before Every Generation

Synthesia supports both scene-level and full-video previews without consuming generation credits. Avatars are not fully animated during these previews, but users can check the narration, scene structure, visual timing, and animations.

Before generating, confirm that:

• The correct project and language are open.
• The script contains no placeholders.
• Scene order is correct.
• Voices and avatars are assigned properly.
• Pronunciations sound natural.
• Visuals match the narration.
• Music remains below the voice.
• Captions and text are readable.
• No private information is visible.

Review the Generated Avatar Performance

After generation, watch the video from beginning to end. This is when the complete avatar animation, gestures, facial movement, and lip synchronization can be evaluated.

Pay attention to emotional sentences, questions, difficult words, long pauses, and transitions between speakers. Edit and regenerate only the sections that genuinely need correction.

Monitor Credits While Editing

Regular video generation currently costs two credits per second on self-service plans. Editing and regenerating a video may use additional credits depending on what new material must be rendered.

To preserve credits:

• Preview before generating.
• Test avatars and voices in short projects.
• Finalize the script before rendering.
• Avoid generating minor unfinished revisions.
• Check the cost of AI-generated media before confirming it.
• Duplicate projects before creating major variations.

Unused self-service credits reset with the relevant billing period rather than accumulating indefinitely.

Duplicate Before Major Changes

Create a copy before changing the language, presenter, audience, format, branding, or overall structure of an approved project.

Keep one master version and name each variation clearly:

• Product Tutorial – Master
• Product Tutorial – New Customers
• Product Tutorial – Spanish
• Product Tutorial – Short Version

This prevents experimental changes from damaging the approved project.

Follow Avatar and Consent Rules

Use personal avatars and cloned voices only with the required permission and consent. Organizations should also define who owns an avatar, who can use it, who approves scripts, and what happens when the person represented leaves the organization.

Do not write stock avatars as though they are real employees, celebrities, customers, or experts providing personal opinions or endorsements. Synthesia may flag scripts that could mislead viewers about an avatar’s identity or role.

Every generated project passes through Synthesia’s content-moderation process, and some material may be routed for manual review.

Review Licensing Before Publishing

Confirm that uploaded music, footage, photographs, logos, and other assets are owned or properly licensed.

Synthesia also applies specific licensing rules to videos using its stock and synthetic avatars or voices. Custom avatars and voices may follow different usage arrangements based on the agreement with the person represented.

Review the current licensing terms before using videos in advertising, paid promotions, endorsements, or other commercial campaigns.

Localize Instead of Only Translating

When creating another language version, review more than the individual words.

Check:

• Regional vocabulary
• Tone and formality
• Date and number formatting
• Units of measurement
• Calls to action
• On-screen screenshots
• Voice accent
• Cultural examples
• Text length and scene timing

A translated sentence may be much longer than the original. Adjust layouts and animations instead of forcing it into the same amount of space.

Keep External Copies

Save important materials outside Synthesia, including the final MP4, script, subtitles, original media, music, translation files, and approval records.

Version history and project storage are useful, but external backups protect finished work when account access, permissions, subscriptions, or workspace organization changes.

The most effective Synthesia workflow is to let AI accelerate repetitive production while the user controls the message, accuracy, pacing, visual clarity, and responsible use of avatars. A simple, well-reviewed video will usually communicate more effectively than a complicated project filled with unnecessary effects.

Frequently Asked Questions

Is Synthesia free to use?

Yes. Synthesia’s Basic plan currently includes 1,200 credits per month, which can create up to approximately 10 minutes of standard video when no credits are used on other features. Standard video generation costs two credits per second.

Do free Synthesia videos contain a watermark?

Yes. Videos created on the Basic plan include the Synthesia logo. Removing it requires upgrading to a paid plan and generating the video again.

Can I download videos on the free plan?

No. The Basic plan does not include video downloads. Users need a Starter, Creator, or eligible Enterprise plan to download the finished MP4.

Free users can still create and watch videos through Synthesia’s hosted platform.

What format does Synthesia use for downloaded videos?

Synthesia exports finished videos as Full HD 1080p MP4 files with a resolution of 1920 × 1080. Users can also download WAV audio, SRT or VTT captions, and XLIFF translation files when those options apply to the project.

Do I need video-editing experience?

No. Synthesia uses a scene-based editor that works more like presentation software than a traditional video timeline. Users can add a script, select an avatar and voice, choose layouts, and arrange visual elements inside the browser.

Beginners should still review the generated script, visuals, pronunciation, captions, and final avatar performance before publishing.

Can I use Synthesia on a phone?

Synthesia provides a mobile-browser experience for self-service users. Mobile users can browse their accounts, watch generated videos, dub content, use AI Playground, create Personal Avatars, and manage plan upgrades.

The complete video editor is not available on mobile. A desktop or laptop is required for full editing functionality. Enterprise mobile access is also not currently supported.

Does Synthesia have a mobile app?

No separate app is required. The supported mobile experience runs through the device’s web browser.

Can I create a video from a PowerPoint?

Yes. Synthesia can import .pptx presentations containing up to 150 slides or a maximum file size of 1 GB. Speaker notes are imported as the video script, while supported text, images, videos, and shapes become editable elements.

PowerPoint animations are not imported and must be recreated using Synthesia’s own animation controls.

Can I create my own AI avatar?

Yes. Eligible users can create a Personal Avatar using a photograph or recorded footage. The person whose appearance is used must complete Synthesia’s identity and consent process. Photo-based Personal Avatar allowances vary by plan.

Users can also work with Synthesia’s included stock avatars without filming or completing an avatar-creation process. Stock-avatar availability depends on the plan.

Can I clone my own voice?

Yes. Voice cloning is currently available on all Synthesia plans. Users can record directly inside the platform or upload an existing voice sample, and the actual speaker must complete the required consent process.

After creation, the voice appears under Custom Voices and can be paired with an available stock or Personal Avatar.

Can several avatars speak in one scene?

Yes. Synthesia’s Dialogue feature supports up to 20 avatars in one scene. Each speaker can be assigned a separate script section and voice.

This can be useful for interviews, role-playing exercises, training scenarios, and customer-service examples.

Can Synthesia translate videos?

Yes. Synthesia supports AI Dubbing for existing videos and provides more advanced video-translation workflows for supported plans.

AI Dubbing can translate spoken audio from an uploaded video or YouTube link. It uses credits on Basic, Starter, and Creator accounts, and optional lip synchronization increases usage.

More advanced localization features, such as connected translations and multilingual publishing, may require an Enterprise plan.

Can I edit a video after generating it?

Yes. Generated videos remain connected to editable projects. Users can reopen a project, change the script, avatar, voice, visuals, or scenes, and generate another version.

Synthesia creates a version whenever the video is generated and also saves automatic project snapshots approximately every 30 minutes. Earlier versions can be restored or duplicated through Version History.

Does editing and regenerating use more credits?

It can. Standard video generation costs two credits per second, and Synthesia recalculates usage when an edited project is generated again. Changes that require new narration, avatars, voices, or video duration may consume additional credits.

Previewing the project does not use standard video-generation credits, so review the script and scenes carefully before generating another version.

Do unused credits roll over?

No. Monthly or annual usage credits reset at the beginning of the applicable billing period and unused credits do not accumulate.

Duplicating a project does not use credits, but generating the duplicated version does.

Can I use Synthesia videos commercially?

Synthesia videos can be used for many business purposes, including training, website product videos, FAQs, YouTube videos, and unpaid social-media content. However, videos using Synthesia stock or synthetic avatars and voices cannot be used in paid advertising or other paid promotional placements under the standard stock-avatar license.

Videos using a custom avatar or voice follow the permission agreement made with the person represented, although Synthesia’s content-moderation rules still apply.

Review Synthesia’s current licensing terms before using any video in paid advertisements, television broadcasts, endorsements, or other promotional campaigns.

Why might Synthesia reject a video?

Synthesia reviews generated content through its moderation system. Stock avatars are intended for neutral, factual, and brand-safe content and may be restricted for sensitive subjects, personal opinions, endorsements, political material, or individualized medical, legal, or financial guidance.

A video containing both stock and Personal Avatars may still be rejected when the stock-avatar sections violate Synthesia’s rules, because moderation decisions can apply to the complete video.

Can I share a Synthesia video without downloading it?

Yes. Generated videos can be shared through a hosted link or embedded on a webpage. Enterprise users can also export supported videos as SCORM packages for compatible learning-management systems.

Sharing through a hosted link can be useful when the project may need future updates because a newer generated version can replace the published version without requiring a completely new distribution process.

Start Using Synthesia Effectively

Synthesia makes it possible to create presenter-led videos without filming actors, recording every voiceover manually, or learning a traditional video editor. Users can begin with a prompt, script, presentation, webpage, document, screen recording, or blank project and then build the video scene by scene.

The strongest results come from preparing the script before spending time on detailed design. Keep sentences natural, divide the information into clear scenes, correct difficult pronunciations, and confirm that each visual supports what the narrator is saying.

Beginners should start with a short practice project. Use one stock avatar, a simple voice, a few scenes, and limited animation. This provides enough experience to learn the editor, preview tools, generation process, and credit system without making the project unnecessarily complicated.

A practical Synthesia workflow is:

1. Define the video’s audience and purpose.
2. Create or import the script.
3. Divide the content into focused scenes.
4. Select the avatar, voice, and language.
5. Correct pronunciation and pacing.
6. Add layouts, media, text, and screen recordings.
7. Apply captions, music, branding, and transitions.
8. Preview each scene and the complete project.
9. Generate and review the finished avatar performance.
10. Share or download the approved version.
11. Save external copies of the video and supporting files.

Synthesia is most useful when its AI handles repetitive production tasks while the user remains responsible for the message, accuracy, visual clarity, and responsible use of avatars and voices.

A simple video with a clear script and relevant visuals will usually perform better than a complicated project filled with unnecessary animation, excessive text, or frequent layout changes. Start with the essential information, review the generated result carefully, and add advanced features only when they improve the viewer’s understanding.

Related Articles

• Synthesia Review (2026): Features, Pricing, Pros & Cons
• How to Use Pictory: A Beginner’s Guide (2026)
• Pictory Review (2026): Features, Pricing, Pros & Cons
• How to Use InVideo AI: A Beginner’s Guide (2026)
• InVideo AI Review (2026): Features, Pricing, Pros & Cons
• Best AI Video Generators in 2026
• Best AI Tools for Content Creators in 2026