HeyGen

The AI video platform that puts a lifelike avatar of you on screen: Avatar V, Video Agent, translation into 175+ languages with lip sync. A free mini-course from first video to pro prompts.

What it nails

• Avatar V from a 15-second phone video
• Translation into 175+ languages with lip sync
• Video Agent builds whole videos from a brief

Watch out for

• Payments are non-refundable

HeyGen promises "AI videos starring you, made in minutes". Record a short clip of yourself once and HeyGen makes a digital twin that can present, teach or sell in 175+ languages. Its newest model, Avatar V, "captures your motion, gestures, and natural expressiveness from a short reference video, then lets you apply all of that to any photo of yourself." Around it sit Video Agent, video translation, voice cloning and live interactive avatars.

Start here: the basics

1. Sign up free: 3 videos a month, up to 1 minute each, with Avatar IV and one custom video avatar.
2. Create your avatar. Avatar V needs a 15-second video you can record on a phone.
3. Write or paste your script (up to 5,000 characters for Avatar IV) and pick a voice.
4. Direct the performance with a short motion prompt, then generate.

Recording tips for a great twin

• "Be extremely expressive - more than feels natural"; "Vary your tone and use your hands."
• "Keep your expression subtle - avoid big smiles or unusual angles".
• "Face too small = lip sync degrades": frame yourself close.
• Record a separate clean track for your voice clone.

Lesson: direct the avatar with motion prompts

HeyGen's rules: "Focus on clear, visible actions: point, nod, turn, wave, smile", "Limit each prompt to one gesture or emotion at a time", and there is "no need to specify seconds". The formula is [Body part] + [Action] + [Emotion or intensity].

Avatar waves with a friendly smile.

Example by HeyGen · source ↗ · Made with Avatar IV custom motion.

Right arm raises in a wave, enthusiastic and friendly.

Example by HeyGen · source ↗ · HeyGen's own example of its formula.

Avatar crosses arms, looking thoughtful.

Example by HeyGen · source ↗

Power moves

1. Video Agent: a finished video from a brief

"The more context and intent you provide, the better the Video Agent can structure scenes, pacing, and visuals." You can also paste a full script and it "will largely follow it scene‑by‑scene".

Use a professional female avatar. Make a compliance training video explaining phishing in detail. Use examples and list top watch-outs. Leverage motion graphics as A-roll overlay and B-roll to help explain core concepts.

Example by HeyGen · source ↗ · An official Video Agent prompt.

Create a 1-minute video about camera aperture. Use minimal science diagrams and visualizations. No avatar needed, only voice-over. Cool neutrals (navy, cyan), thin-line diagrams, and slow elegant motion.

Example by HeyGen · source ↗

Use #1E40AF as primary blue, #F8FAFC as background white, and #0F172A for text.

Example by HeyGen · source ↗ · Add this line to any Video Agent prompt.

Refine my prompt for HeyGen Video Agent. Keep my original intent, but rewrite it into clear, structured production instructions optimized for Video Agent.

Example by HeyGen · source ↗ · HeyGen's own prompt-refiner, to paste into any chatbot with your rough idea.

2. Translate yourself into 175+ languages

"Translate videos into 175+ languages with realistic AI voices, lip sync, and subtitles", keeping your own voice through voice cloning, and "translate a single video into up to 10 languages at once."

3. Cinematic scenes with Seedance 2.0

HeyGen added Seedance 2.0 in April 2026. Its guide's structure: "Subject → Action → Environment → Camera → Style → Constraints", and "Focus your prompt on motion".

A woman with her hair pulled into a sleek low bun, wearing an oversized black wool blazer over a silk slip dress, walks confidently across a marble gallery floor. Dramatic side lighting carves strong shadows across her face, high contrast editorial mood. Slow push-in toward her face as she turns to camera. Cinematic, shallow depth of field, 35mm film grain. Smooth motion, stable framing.

Example by HeyGen · source ↗ · From HeyGen's Seedance 2.0 guide. The Lab runs Seedance 2.0 too: this button opens it in our Video studio.

A woman wearing a cream blouse and tailored beige trousers, stands confidently in a sunlit rooftop garden. Golden hour lighting, warm and soft, romantic glow on her face and hair. Low angle hero shot looking up at her, projecting authority. Cinematic, shallow depth of field. Smooth motion, stable framing.

Example by HeyGen · source ↗ · From the same guide. Try it in the Lab with Seedance 2.0.

4. Live, interactive avatars

LiveAvatar and the Avatar Realtime API let your avatar talk with viewers in real time, for support, sales or training.

Plans and credits

• Costs: Avatar IV photo look 16 credits a minute, video look 31; Avatar V 48 credits a minute.
• "Monthly subscribers' unused credits roll over for one more month. Annual subscribers' credits accumulate until their annual renewal date." They do not carry over after cancelling.
• "Payments are non-refundable, except where required by law."

Make the pieces in the Lab

AI Whisperer Lab does not make talking avatars. But many videos need the parts around them, and those run in the Lab by the run, with credits that never expire:

• Voiceovers with ElevenLabs, the same voice engine HeyGen offers: text to speech, voice cloning and speech to speech in the Audio studio.
• Cinematic b-roll with Seedance 2.0 and 2.5, Kling 3.0 or Veo 3.1, which makes "video with native audio and speech", in the Video studio.
• Thumbnails and scenes with Nano Banana Pro or GPT Image 2 in the Image studio.

FAQ

Sources: HeyGen · Avatar V · HeyGen pricing · Motion prompts · Video Agent prompt guide · Moderation policy