Introducing H3 Max, new post-trained video model by fal Research.
H3 Max ranks #1 for overall quality, prompt understanding, and aesthetics against leading video models, on both first-party and third-party independent evaluations while generating a 5-second 720p video under 3
Reference-to-video on H3 Max is now generally available. Up to 2X faster, with significantly higher quality.
-> RTF down from 1.5 to 0.876, so real-time generation now works with up to 4 references
-> Better context understanding: references and prompt are automatically aligned
Reference-to-Video is now live for MiniMax H3 Max on fal!
This is an early preview that can achieve up to RTF=1 (and with references around RTF=1.5) on 768p, and we aim to improve the speed by 2x while slashing costs during this week. Quality wise it delivers the same
The early preview of H3 Max R2V (reference-driven generation) is #1 for realistic video style transfer, with a 73.9% win rate. Doing this while delivering sub-1 RTF.
The frontier video model is also setting a new bar for video editing.
Try it yourself today
Realistic video editing has a new leader on style transfer.
MiniMax H3 Max Video Edit moves to 2nd of 11 on realistic, up from 4th for base MiniMax H3, and posts the highest win rate on the realistic board:
- 1st, style transfer (restyling a shot while keeping its motion) at
We're excited to have Jonathan Gur-Zeev @Ygurzeev, Director of Product at Figma Weave, speaking at GMC 2026.
Jonathan spent years working in post-production before co-founding Weavy, which later became a part of Figma. He’ll talk about what sets a product apart when every team
