A cheap way for developers to get short clips with sound through fal, with five free generations a day to test it.
Kenning's take
What it does
H3 Max is a video model that fal built by post-training MiniMax's open-weight H3 model for better prompt adherence and look. It makes 5 to 15 second clips at 480p or 768p with sound generated alongside the picture. It runs on fal through text-to-video, image-to-video and reference-to-video endpoints.
- Text-to-video, image-to-video and reference-to-video
- Clips of 5 to 15 seconds at 480p or 768p
- Synchronised audio in the same pass
- Reference-to-video with up to 12 files
- Five free generations a day on the tool page
Who it suits. Developers and video makers who want fast, low-cost AI video through an API.
Key facts
- Category
- AI models
- Part of
- Hailuo AI
- Made by
- fal (post-trained from MiniMax H3)
- Based in
- United States
- Launched
- 2026
- How to get it
- through fal's API and playground; five free generations a day on the tool page
- Works on
- Web, API
- Free plan
- five free generations a day on the fal tool page
- API price
- $0.05 per second at 480p, $0.08 per second at 768p (after launch discount ended 1 September)
Pricing
| Plan | Price | Billed |
|---|---|---|
| API 480p | $0.05 | second of video |
| API 768p | $0.08 | second of video |
FAQ
Who made H3 Max?
fal, by post-training MiniMax's open-weight H3 video model; fal says it was released jointly with MiniMax.
How much does H3 Max cost?
On fal it costs $0.05 a second at 480p and $0.08 a second at 768p, with five free generations a day on the tool page.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools
Hailuo 2.3
MiniMax's video model with better body movement, facial expression and art styles
ACE-Step 1.5
Open-source music generation model that writes full songs with lyrics and runs on ordinary consumer GPUs.
AlphaGenome Atlas
Google DeepMind database of AI-predicted effects for every possible single-letter change in human DNA.
Amazon Nova 2
Amazon's second family of Nova AI models, covering reasoning, speech conversation and combined text-and-image generation.