The leading text-to-video model

HeyGen Video is HeyGen's general-purpose AI video model, built on MiniMax H3 and post-trained by HeyGen. It generates an entire scene from a prompt, including the subject, setting, lighting and sound, as a 5- to 15-second clip. It works in three modes: text to video, image to video and reference to video. On HeyGen, you can use it alongside your AI digital twin and other models within a single workflow.

You can create brand film shots, product scenes, B-roll cutaways, social media hooks and stylised explainers, each with matching audio. HeyGen Video maintains consistency across scenes: objects retain their size and position while following the laws of physics. It follows a brief precisely, so dialogue is included only when requested. It works best for short, self-contained shots featuring one subject, one location and one action.

HeyGen Video is available through the HeyGen API as the heygen-video-1 model with paid API keys. Create a key in your API settings, then send a prompt and mode to the create endpoint. Completed clips will appear in your workspace alongside your other HeyGen videos.

Choose a mode, write a detailed prompt and submit it. Use text to video for a new scene, image to video to begin with your own first frame, or reference to video to retain a product, person or place that you provide. Detailed prompts specifying the camera, lens and sound work better than short ones. Poll the job or set a callback, then download the MP4.

Clips can be between 5 and 15 seconds long at 480p, 768p, 1080p or 2K, encoded as 24 fps MP4 files with stereo audio. Supported aspect ratios include 21:9, 16:9, 4:3, 1:1, 3:4 and 9:16, allowing you to create content for YouTube, TikTok, Reels or Shorts without cropping it again. Videos in 1080p and 2K render only in 16:9 or 9:16 and are billed at three times the 768p rate.

Yes. Videos generated with HeyGen Video can be used for commercial purposes, including advertisements, social media posts, product launches and client deliverables, under your HeyGen commercial licence. To keep fictional branding out of the frame, add "no logos, brand names, printed words or badges" to your prompt, as plain clothing and equipment may otherwise feature fabricated marks.

Yes. Room tone, sound effects and speech are generated along with the visuals from the same prompt, ensuring the audio matches the setting you described. Written dialogue works best at roughly 2.5 words per second, or about 37 words in a 15-second clip. Specify "no music" if you do not want background music. For longer scripted content, your digital twin can speak in your own voice.

Seedance 2.0 and Veo 3.1 are known for complex, multi-shot cinematic motion. HeyGen Video stands out for its control: precise prompt adherence, sound generated from the same prompt, support for up to 12 references and repeatable seeds. HeyGen's greater advantage lies in its workflow. You can choose the right model for each shot, then add avatars, voice and captions—all in one place.

Yes. Set the seed yourself, and the same prompt with the same seed, submitted consecutively, returns an identical file. Keep the seed unchanged and modify one clause at a time to iterate. When the seed is omitted, the server selects one at random, so two identical requests return different videos. The completed job reports the seed it used.

Only if you allow it. Prompt enhancement can reword the rest of the prompt, so disable it when you want your text retained exactly as written. You can also provide a callback URL to be notified when a job is completed, so you do not have to poll for the result.