Give your script a face — pre-recorded or live
AI avatars turn a script and a face (plus a voice) into video of someone talking — either rendered once ahead of time, or generated live in a real conversation. Pre-recorded avatars (HeyGen, Synthesia, D-ID) are the mature, polished end of the market: upload a photo, type a script, get a video in minutes. Realtime interactive avatars (Tavus, and HeyGen/D-ID live products) go further, holding an actual back-and-forth conversation on camera with an LLM behind the face. On the open-source side, SadTalker and LatentSync/Wav2Lip cover photo-animation and lip-sync dubbing for free, if you have a GPU. Be honest about the gap: there is no production-ready open-source option yet for realtime interactive avatars — if you need live conversation today, you are choosing between paid proprietary tools. → Quick pick: polished pre-recorded talking-head → HeyGen or Synthesia; a photo into a talking head → D-ID; live interactive conversation → Tavus; free with a GPU → SadTalker (pre-recorded) or LatentSync (lip-sync).
HeyGen and Synthesia are the polished pre‑recorded avatar services – upload a photo, type your script, and receive a finished talking‑head video in minutes.
SadTalker lets you animate a static photo into a talking head, while LatentSync with Wav2Lip provides lip‑sync dubbing – both run locally on a GPU and are free to use.
HeyGen’s Photo Avatar feature that turns a single image into an unlimited talking video.
You will have generated a short video of a realistic avatar speaking your script from just one photo.
A browser‑based feature that creates a speaking video from a single portrait by analysing facial geometry and syncing lip movements to supplied audio or text.
Generate a short avatar video that lip‑syncs to your script using only an uploaded photo and optional voice input
HeyGen’s Motion Engine, which applies preset gesture and facial dynamics to an existing avatar video.
Add coordinated body language that matches a chosen emotional tone to an existing avatar video
HeyGen’s Avatar IV motion controls that let you customize facial expression, gesture, and upper‑body animation.
You will enhance the previously created avatar video with natural gestures and expressive face dynamics.
An AI‑driven translation tool that replaces an avatar video's audio track with multilingual speech while keeping the original visuals unchanged.
Create a version of the avatar video that speaks a different language while preserving all visual elements
HeyGen’s Video Translation feature that automatically translates speech, subtitles, and lip‑sync to any of 175+ languages.
You will produce a version of your expressive avatar video in a different language with accurate lip‑sync.
HeyGen’s export and sharing workflow that lets you download high‑resolution files or copy a shareable link.
You will make both the original and translated videos publicly accessible for review or distribution.
43 outcomes in all — one per recipe below.
Record a short 15‑second clip with clear lighting and natural speech, then go to Avatar → New Avatar → Clone a real person (or Digital Clone). Upload the video, optionally add a voice sample, and let HeyGen process it into a digital twin that mimics your appearance, facial movements and voice. Once created you can save the avatar for any future project.
In AI Studio click the Motion Engine icon and choose a preset energy style such as excited, welcoming or calm. The engine automatically animates gestures and facial dynamics to match the tone, so you get coordinated movement without manual keyframing. Preview the result and regenerate the video to apply the changes.
Yes. From the home page click Translate, select Video Dubbing, upload your original MP4 and choose the target language. HeyGen translates the script, generates a new voice in that language and syncs it with the existing visuals, producing a dubbed version you can preview and download.
Use Avatar Shots: go to Avatar → Avatar Shots, select your digital clone, and write a concise prompt describing the scene, setting and outfit changes. Choose portrait or landscape orientation and click Generate; HeyGen synthesizes rapid camera moves and wardrobe swaps into a 15‑second cinematic promo.
Record a short 15‑second clip with good lighting and natural speech, then upload it in HeyGen’s Avatar 5 or Digital Clone creator. The platform extracts your facial movements, voice timbre and speaking style to generate a personalized avatar that mimics your look and voice.
Use the Motion Engine inside AI Studio: click its icon, choose a preset energy style such as excited, welcoming or calm, preview the movement and regenerate the video. The engine automatically animates facial dynamics and gestures without manual keyframing.
Yes. From the home page click Translate, select Video Dubbing, upload your original MP4, choose the target language and start the process. HeyGen will generate a new voice in that language and sync it with the existing visuals.
Open Video Agent, paste or type your script into the prompt box, then select the avatar and voice you have saved. Choose a visual style and any brand assets, click Generate (or Auto‑Generate) and HeyGen will auto‑storyboard scenes, apply backgrounds and render the final video.
Only a 15‑second clip and need a digital double
HeyGen can clone a real person by recording just 15 seconds of video. The system extracts facial movements, gestures and expressions to build an avatar that mimics you across angles and scenes.
Need new outfits and settings for your avatar
Design with AI lets you define a base portrait and then either pick from pre‑made scenes or write prompts to create new outfits, backgrounds and angles. The avatar’s face stays consistent while the surrounding visuals change.
I have a text script and want an avatar to speak it
AI Studio combines a text script (typed or generated via ChatGPT) with an avatar, motion engine, and chosen look to output a finished video in minutes. Advanced settings let you control motion style for more natural movement.
Want a multi‑scene video without editing
Video Agent automates script writing, scene layout and avatar rendering. By describing topic, audience and tone in one prompt, the system creates a full video with multiple scenes without manual editing.
Need a digital double for videos
Synthesia lets you upload photos and voice recordings to generate an avatar that mirrors your appearance and speech. The platform trains the model on your data, producing a lifelike representation you can use in videos.
Need a video narrator that shows feeling
Synthesia offers a library of over 200 avatars with selectable expressive styles, allowing you to match tone and body language to your script for more engaging content.
Need to make videos of yourself without filming
Synthesia lets you generate an avatar that looks and sounds like you by uploading a single photo or a short video. The platform processes the media to build a 3D model with voice synthesis, enabling you to produce videos without ever filming again.
Want your avatar to match your brand?
Within the editor you can swap an avatar’s clothing and location by describing what you want. Synthesia’s generative AI creates a preview in real time, letting you match branding or context without external design tools.
Want one video that reaches every market
After finalizing a video, Synthesia can automatically translate both spoken audio and on‑screen text into over 140 languages, producing localized versions in one click. This streamlines global distribution without manual dubbing or subtitle creation.
By selecting the avatar and using the Media tab, you can describe any desired action (e.g., ‘raise hand’, ‘point to chart’). Synthesia’s AI interprets the prompt and animates the avatar accordingly, adding dynamic movement without manual keyframing.
The video defines six essential words (Avatar, Voice, Project, Scene, AI Studio, Credits) that appear throughout HeyGen. Knowing these terms lets you recognize UI elements and plan your videos without getting lost when labels change.
The left‑hand sidebar contains five main sections (Home, Avatar, Brand, Apps, Projects). Understanding this layout lets you jump directly to the tool you need, regardless of UI updates.
Need a fast prototype video from just a text prompt
Quick Create is a shortcut that generates a video from a simple text prompt, bypassing the full editor. It’s useful for rapid prototyping or testing ideas before committing to detailed editing.
Want every video to match my brand colors and logo
The Brand section stores colors, logos, fonts, and templates. By configuring a brand kit once, all future videos can automatically use the same styling, saving time and ensuring consistency.
Need a custom talking character for a video
D-ID Studio lets you build a video featuring a custom AI avatar by selecting or creating an avatar, adding a script, choosing or cloning a voice, and layering visual elements. The platform integrates avatar generation, text‑to‑speech, and scene design in one workflow, making high‑quality avatar videos fast and accessible.
Need a presenter that matches your audience
HeyGen offers a searchable library of pre‑made avatars filtered by age, gender and ethnicity. Selecting the right one saves time and ensures your video matches its target audience.
Want a digital double of yourself
By recording a short 15‑second clip with clear lighting and natural speech, HeyGen learns your facial movements, voice timbre and speaking style, producing a personalized AI clone.
Need a custom animated avatar for video
Using the “Design with AI” option lets you describe a character’s appearance, age, gender, ethnicity and style; HeyGen generates multiple visual variations that can be saved and paired with any voice library.
Video script, captions or background look off
The AI Studio interface separates script, timeline, avatar/media/captions controls, allowing precise tweaks to wording, caption styling and background images generated on‑the‑fly with AI image models.
Want expressive avatar gestures without keyframing
The Motion Engine lets you pick a preset energy style (excited, welcoming, calm) that automatically animates gestures and facial dynamics, giving the avatar more life without manual keyframing.
Need the same avatar video in another language
HeyGen can automatically translate the script, generate a new voice in the target language and sync it with existing visuals, letting you reach multilingual audiences without re‑recording.
Want a portrait that talks
The tool lets you upload any portrait and animates it to speak using AI-generated facial movements. It works by mapping the static image onto a 3D model that syncs lip movements with audio, producing realistic talking heads.
Can’t record audio for your avatar
D‑ID provides a library of synthetic voices that can be matched to any avatar. By selecting a voice and adjusting style, you control tone, accent, and pacing without recording audio yourself.
Need a custom video presenter but have no artwork
Beyond pre‑made presenters, D‑ID can create new avatar images on the fly using a text prompt that feeds a Stable Diffusion model. This lets you design custom characters without external image tools.
Need a video character in a specific outfit and background
Synthesia lets you change an avatar's look by describing the desired costume or setting. The platform generates a preview instantly, allowing rapid iteration without any design skills.
Want a personal presenter that looks and sounds like you
By recording a short video of yourself reading a script, Synthesia extracts facial movements and voice characteristics to build an avatar that looks and sounds like you, enabling fully automated personal presentations.
Want a video version of yourself
You record 15 seconds of video and optionally a voice sample, then HeyGen generates a digital twin that mimics your appearance, speech, and motion. The short clip captures enough facial and body cues for Avatar 5 to produce authentic movements.
Only a script in mind
Video Agent lets you describe the entire video (topic, style, length) in natural language; HeyGen assembles scenes, adds your avatar, and renders the final clip automatically. This bypasses manual editing for quick content production.
Want a fast, cinematic promo for your avatar
Avatar Shots uses Seedance 2.0 to synthesize rapid scene changes, wardrobe swaps, and camera moves based on a single prompt, producing high‑impact short clips without filming.
Need your video in another language
The translation feature re‑uses your avatar’s visual content while swapping the audio track with AI‑generated speech in the target language, preserving brand consistency across markets.
Need a video with you in any clothes and voice
Synthesia lets you modify an avatar’s outfit, background, and even upload a custom likeness that mimics your voice and appearance. By describing the desired look in plain text, the platform generates a preview instantly, allowing rapid iteration without any video editing skills.
Beyond static speaking, Synthesia can prompt an avatar to execute actions like walking, cooking, or demonstrating objects by adding a simple textual cue. The system maps the cue to motion libraries, creating dynamic, engaging video segments without manual animation.
Want a custom video character
HeyGen’s Auto Avatar feature lets you build an avatar either from a public library or by uploading reference images to generate a digital twin. The system analyses facial features and maps them onto a 3D model, enabling realistic lip‑sync and expressions.
Want your AI avatar to speak with your own voice
AutoVoice can clone an existing voice by analyzing a short audio sample, then synthesize speech that matches tone and cadence. This enables the avatar to speak with a natural‑sounding, personalized voice.
Need a video of a custom avatar speaking your script
By entering a text prompt or full script, HeyGen’s Video Agent auto‑storyboards scenes, matches them with the selected avatar and voice, applies visual styles, and renders the final video in minutes.
Need a video presenter that looks and sounds like you
You record a short video of your face and voice, then Synthesia processes it into a personal avatar that mimics both appearance and speech. This works because the platform uses facial mapping and voice cloning to generate a realistic digital presenter.
Need a corporate‑look AI presenter
You can adjust shirt color, add a logo, and select background scenes for ready‑made avatars, ensuring visual consistency with your corporate identity. The platform applies these style changes instantly across the avatar’s appearance.
Want different viewpoints in one video
Recording three personal avatars from different camera angles lets Synthesia switch viewpoints within a single video, creating a more engaging and professional look. Each angle is treated as a separate avatar that can be swapped scene‑by‑scene.
Need your own avatar as presenter in a video
After generating a video outline, you can replace the default presenter with your personal avatar, adjust its size/position, and change background settings. This lets you personalize every scene without re‑rendering the whole script.
Want a video double of yourself
Record a short, well‑lit video of yourself speaking naturally for about 15 seconds. HeyGen extracts your voice, facial expressions and gestures to build an avatar that can be reused in any scene.
Got a script and need an edit‑ready video
Video Agent first builds a scene‑by‑scene blueprint, letting you choose avatar, voice, background and style before any rendering occurs, so you can edit each element without re‑generating the whole video.
Long video or podcast needs bite‑size social clips
Upload a full‑length video or podcast; HeyGen analyzes speech and visual cues to identify key statements and emotional peaks, then outputs trimmed, formatted snippets ready for social platforms.
Only have a headshot?
The system infers facial geometry and natural movement patterns from one image, then animates it to deliver any script, useful for mascots or quick spokespersons when full video isn’t available.
The same set on /recipes, filtered by tool and role.
Shows how to edit scripts, add captions and backgrounds in HeyGen AI Studio, pick a public avatar, and create a custom animated AI avatar
Shows how to generate a speaking avatar from a photo, build a realistic digital twin from a short clip, and convert a script into an editable video using HeyGen tools
Shows how to use HeyGen’s translation, digital clone, and video agent features
Shows how to start using HeyGen by launching a quick video, explaining key terminology, and navigating the sidebar
Upload a portrait (preferably neutral) in Creative Reality Studio, add your script text, choose language, voice and style, then click Generate. The platform maps the image onto a 3D model that syncs lip movements to the audio and produces an MP4 you can download.
Yes. After uploading your image, go to the Voice section, browse the synthetic voice library, preview each option, pick one, optionally adjust its style (e.g., excited or friendly), enter your script and generate the video.
D‑ID can generate a new avatar image using a text prompt that runs through a Stable Diffusion model. In the studio, click “Generate AI presenters,” type a descriptive prompt, select resolution, and submit; the result is a fresh avatar ready for script and voice addition.
Record a short 15‑second clip with clear lighting and natural speech, then go to Avatar → New Avatar → Clone a real person (or Digital Clone). Upload the video, optionally add a voice sample, and let HeyGen process it into a digital twin that mimics your appearance, facial movements and voice. Once created you can save the avatar for any future project.
In AI Studio click the Motion Engine icon and choose a preset energy style such as excited, welcoming or calm. The engine automatically animates gestures and facial dynamics to match the tone, so you get coordinated movement without manual keyframing. Preview the result and regenerate the video to apply the changes.
Yes. From the home page click Translate, select Video Dubbing, upload your original MP4 and choose the target language. HeyGen translates the script, generates a new voice in that language and syncs it with the existing visuals, producing a dubbed version you can preview and download.
Use Avatar Shots: go to Avatar → Avatar Shots, select your digital clone, and write a concise prompt describing the scene, setting and outfit changes. Choose portrait or landscape orientation and click Generate; HeyGen synthesizes rapid camera moves and wardrobe swaps into a 15‑second cinematic promo.
Record a short 15‑second clip with good lighting and natural speech, then upload it in HeyGen’s Avatar 5 or Digital Clone creator. The platform extracts your facial movements, voice timbre and speaking style to generate a personalized avatar that mimics your look and voice.
Use the Motion Engine inside AI Studio: click its icon, choose a preset energy style such as excited, welcoming or calm, preview the movement and regenerate the video. The engine automatically animates facial dynamics and gestures without manual keyframing.
Yes. From the home page click Translate, select Video Dubbing, upload your original MP4, choose the target language and start the process. HeyGen will generate a new voice in that language and sync it with the existing visuals.
Open Video Agent, paste or type your script into the prompt box, then select the avatar and voice you have saved. Choose a visual style and any brand assets, click Generate (or Auto‑Generate) and HeyGen will auto‑storyboard scenes, apply backgrounds and render the final video.
Record a short, well‑lit 15‑second video of yourself speaking naturally and upload it via the Avatar 5 or Digital Clone creator. The system extracts your facial movements, voice timbre and speech style to build a digital twin. After processing you can review the generated avatar, select the best version and save it for any future project.
Yes, use the Motion Engine in AI Studio. Choose a preset energy style such as excited, welcoming, or calm, which automatically applies coordinated facial dynamics and hand gestures. Preview the motion, adjust if needed, then regenerate the video to apply the changes.
From the home page click Translate, select Video Dubbing, and upload your original MP4. Choose the target language; HeyGen will translate the script, generate a new voice in that language, and sync it with the existing visuals. When processing finishes you can preview and download the dubbed version.
Use Quick Create from the Home screen: enter a brief text prompt or script, pick a stock avatar if needed, and submit. HeyGen generates a one‑minute prototype video automatically, which then appears under Projects for you to review or edit further.
AvatarAI StudioMotion EngineVideo AgentAvatar ShotsSeedance 2.0Nano BananaAutoAvatarAutoVoiceCreditsBrand KitQuick CreatePersonal AvatarAvatars → Customizehex codeEnterprise planAvatar dropdownChange AllAdd SpaceCustom AvatarAction PromptMedia tabTranslate (in export)Multi‑Angle AvatarsAsk, share, or report — over on the Heidelberg AI community forum.