Slop TVNewsLatest
News

HeyGen's first video model sells from 1 cent a second

HeyGen Video is a fine-tune of MiniMax's H3 model and generates 5 to 15 second clips with their own audio, at half its regular price through October.

Illustration: HeyGen's first video model sells from 1 cent a second
Illustration: SLOP TV News. Logos are trademarks of their owners.

Key takeaways

  • HeyGen Video is the company's first general-purpose video model, post-trained on top of MiniMax's open H3 model.
  • It generates 5 to 15 second clips with their own dialogue, ambience and sound effects in one call, so no separate speech or lip-sync pass is needed.
  • Launch pricing is 1 cent per second at 480p and about 1.5 cents at 768p, half the regular rate, through the end of October.
  • Access is API-first through the model heygen-video-1, with ComfyUI and a web interface planned to follow.

HeyGen has shipped its first general-purpose video model, and it is a fine-tune rather than a model trained from scratch. HeyGen Video is built on MiniMax's open H3 model and post-trained by HeyGen, according to the company's developer documentation. The company made its name with avatar video, where a script drives a presenter you already own; this model invents the subject, setting, light and sound from a prompt instead.

One model, heygen-video-1, is called three ways. Text to video takes a prompt. Image to video takes a prompt and one image, which becomes the clip's first frame. Reference to video takes a prompt plus your own images, video clips or audio and keeps the products, people or places you supplied. Each generation runs 5 to 15 seconds and returns its own audio track of dialogue, ambience and sound effects in the same pass, so there is no separate speech step, lip-sync stage or avatar to select.

The pitch is price. MindStudio reports a regular rate of 2 cents per second for 480p output and about 3 cents for 768p, with a launch sale through the end of October halving both: 1 cent per second at 480p and roughly 1.5 cents at 768p. Access is API-first, with ComfyUI integration and a native web interface described as later additions rather than available at launch.

That pricing is the story for anyone who has watched per-second rates this year. Google's Veo 3.1 tiers run from $0.05 per second for Lite at 720p up to $0.60 for the full model at 4K, and a host of smaller vendors have been advertising single-digit cents. HeyGen is not claiming a new quality ceiling: the base model is H3, which already sits under other products, so its motion and prompt handling should resemble other H3 fine-tunes, and the differentiation is cost and distribution.

For creators, a cent a second at 480p is a useful preview tier rather than a delivery format. The real number to watch is what 768p costs once the sale ends, because that is the resolution a social cut is usually finished at, and it doubles.

The API is live now; HeyGen's docs list ComfyUI and a web interface as planned, and the halved rate runs through the end of October.

Sources

  1. developers.heygen.com - the model's mode list, clip lengths and audio behaviour
  2. mindstudio.ai - launch and regular per-second pricing and the H3 lineage
  3. minimax.io - the base model's own specifications
  4. developers.heygen.com - the API catalogue entry
  5. ai.google.dev - Veo 3.1 per-second rates by resolution