Meet HeyGen Video, our general-purpose video model. Describe a scene and get picture and sound together in one call, via text, image or reference to video.
Introducing HeyGen Video, our general-purpose video model. One prompt generates the whole scene: subject, setting, light and sound, in a single call. Create 5- to 15-second clips from text, images or your own references, in styles from live action to stop-motion, at up to 2K. Available now via the HeyGen API.
Most AI video is still built in pieces. One model makes the picture, another voices the lines, lip-sync stitches them together and a sound library fills the gaps. The result: room tone that doesn't match the room, footsteps that land a beat late.
We built a model without the seams. With HeyGen Video, picture and sound come from the same pass, so they fit from the first frame.
What is HeyGen Video?
HeyGen Video is our general-purpose video model. Just describe the scene you want, and it creates the footage and its matching soundtrack together, all from the same prompt.
Every clip runs 5 to 15 seconds and ships with its own audio track: dialogue, ambience and sound effects, generated alongside the image. There is no separate speech pass, no lip-sync step and no avatar to pick first.
That makes it a different tool from the avatar engines HeyGen is known for. Avatar V animates a look you already own from a script you already wrote. HeyGen Video invents what isn't there yet: the subject, the setting, the light and the sound, all from a description.
One model, heygen-video-1. Three ways to call it.
How Does HeyGen Video Work?
Text to video
Start from nothing but words. Describe the subject, the place, the camera and what you want to hear, and HeyGen Video builds the shot from scratch.
Image to video
Hand it a still and it becomes frame one. The clip opens exactly as your image looks and moves forward from there, which makes it the natural way to bring a key visual, a storyboard frame or the last frame of a previous shot to life.
Reference to video
This is where it gets practical for brands. Supply up to 9 images, 3 videos and 3 audio files (12 in total), then address them in your prompt by name, such as <Picture 1>. Give it a photograph of your product and its colour, form and markings survive into a scene you describe around it.
Direct it like a cinematographer
HeyGen Video rewards detail. Prompts of a few hundred to a few thousand characters produce materially better output, and there's no penalty for being specific. A few habits make the biggest difference:
- Name the camera, not the mood. "Cinematic" does almost nothing. A camera body, lens and aperture change the depth of field, grain and motion all at once.
- Ask for texture, then rule out the fix. Request real skin and materials, then add "no retouching, no beauty filter" so faces don't drift toward a polished look.
- Direct the sound. Name the room tone, the effects and how far away they are. Say "no music" if you don't want a score.
- Lock the seed to iterate. The same prompt and seed return the same clip, so you can change one clause at a time and see exactly what it did.
How HeyGen Video Differs From Other HeyGen Products
Most video models can produce one striking frame. The problem shows up a few seconds later: the cup that changes size, the chair that wasn't there a second ago, the voice that sounds like a different room. HeyGen Video was post-trained for the shot as a whole.
Scenes that hold together
Objects keep their size and shape, stay where they were placed and obey physics. Nothing duplicates, vanishes or reappears halfway through.
It does what you asked
HeyGen Video follows a brief literally instead of improvising around it. If you didn't ask for dialogue, you don't get dialogue.
Picture and sound, born together
Room tone, effects and speech come from the same prompt as the image, so they match the space you described. A train arriving on an empty platform sounds like a train arriving on an empty platform: near silence, then the horn, the wheels on the rail joints, the rush of air and the brakes, in that order.
Your product, kept intact
Give it a photo in reference to video and the object survives into the shot with its colour, form and markings intact. That's the difference between a generic product shot and your product shot.
It shines brightest on short, contained shots: one subject, one place, one action and a locked or gently moving camera. That happens to describe most of the shots brands need.
Any style you can name
In HeyGen Video, style is just another part of the prompt. Our launch film shot the same driver close-up in four different media, then took a single drawing from pencil to oil paint to a clay model, each step starting from the last frame of the one before.
The trick is to name a real production process and the marks it leaves, rather than a vibe. Some of the styles in the launch film, and what teams use them for:
Titles that live inside the shot
Every title in the HeyGen Video launch film was generated in camera: painted on a race track, stamped in ink, pressed into cotton paper by letterpress. Short phrases render reliably when the prompt spells them out exactly and names them as the only lettering in the frame. For longer copy, like a full lockup with a tagline, add real type in the edit.
HeyGen Video is available today through the HeyGen API and Creative Studio.
One request starts a job, and you poll or take a callback for the finished MP4.
Each job can also be read through GET /v3/videos/{video_id} alongside the rest of your HeyGen videos, so it slots into pipelines you've already built. HeyGen Video runs on paid API keys.
Pair it with Avatar V
Avatar V is you. HeyGen Video is everything around you.
Avatar V delivers your message with the stable, long-form presence that professional video needs, from a 15-second recording. HeyGen Video fills in the shots no avatar can: the cold open on an empty platform at dawn, the product turning on a wooden table, the B-roll that cuts away while you keep talking.
Together they cover the whole edit. Open on a HeyGen Video scene that earns the watch, hand off to Avatar V to deliver the message, and cut back to a reference-to-video shot of your actual product to close. One platform, one API, one video.
What comes next
The best video tools disappear into the work. You stop thinking about which model drew the picture and which one made the sound, and start thinking about the shot.
That's what HeyGen Video is for. Describe it, hear it, ship it.
Read the HeyGen Video docs · Get an API key · See the launch film, prompt by prompt
Greetings! My name is Ayesha Shaheryar. My words have helped millions over the past two years. As a HeyGen expert and a writer, I am here to introduce tips and tricks to edit your next video in no time.
