Published: October 26, 2026
Video Model
HeyGen Video 1.0 is HeyGen’s first general-purpose video model, built on MiniMax H3 and post-trained by HeyGen. It makes 5 to 15 second clips from a prompt or a start frame, with dialogue, ambience and effects in the same pass.
In October 2026 it ranked first on Artificial Analysis’ blind text-to-video leaderboard (silent), just ahead of Wan 3.0 and MiniMax H3. It is strongest on single, contained shots with a locked or slow camera: presenters, products and short social clips.
Variants at a Glance
Type | Modalities | Max Res. | Talking Points | |
|---|---|---|---|---|
| Variant | ||||
| Text to Video | Video | Text | 2K | Prompt only, 16:9 or 9:16, 480p to 2K. |
| Image to Video | Video | Text + start frame | 2K | Animates a start frame; the image sets the shape. |
Key Features
Video Generation
- Text-to-Video: A full clip from a written brief, in landscape or portrait.
- Start Frame: Upload a still and it becomes the first frame; look, wardrobe and lighting carry through.
Sound and Prompting
- Generated audio: Every clip comes with sound. Quoted lines are spoken with lip sync, plus ambience and effects.
- Prompt Enhancement: Off, Turbo or Quality. Turbo fills gaps in short prompts; Off sends a finished brief as written.
Technical Capabilities
Details | |
|---|---|
| Quality / Resolution | 480p, 768p, 1080p, 2K |
| Duration | 5 to 15 sec, any whole second |
| Aspect Ratios | Text to Video: 16:9, 9:16. Image to Video: follows the uploaded image |
| Input Modalities | Text prompt; optional start frame image (up to 16 MB) |
| Audio | Always on: dialogue, ambience, effects, optional music |
Limitations
- Start frame is the only image input in Artlist: no end frame, reference images or video.
- Text to Video offers 16:9 and 9:16 only; other shapes need a start frame in that shape.
- Hands operating objects, loose hair and fabric are the least reliable parts of a shot.
- Faces drift toward retouched skin and plain surfaces can pick up invented logos unless the prompt rules them out.
- Unspecified sound is the model’s choice and may add music; Prompt Enhancement can change details of a full brief.
Prompting Tips
- Name the camera, lens and support, and say when the camera stays still. “Cinematic” does little.
- Budget about 2.5 spoken words per second (about 12 words in 5 sec); quote the line and attach it to a described person.
- End with an “Audio:” line placing sounds close, mid and far, in order. Write “No music” to rule it out.
- Exclude retouching, beauty filters and logos in commercial prompts.
- With a start frame, spend the prompt on motion, camera and sound; a closed-mouth portrait starts cleaner.
- Turn Prompt Enhancement off once your prompt is a complete brief.