fal's H3 Max: RL post-training cuts 5s video generation from 120s to under 3s
isidentical · x · 2026-09-16
fal Research launched H3 Max, a post-trained video model claiming #1 in overall quality, prompt understanding, and aesthetics against leading video models on first- and third-party evals, while generating a 5-second 720p video in under 3 seconds (50% off for a week).
Head of engineering isidentical explained the approach at MTS:
- Built on the open-source MiniMax H3 video base model, with reinforcement learning applied to verifiable concepts like video editing and prompt adherence
- Generation time cut from 120s to under 3s — nearly 2x real time — while improving quality over the open-source version and many frontier closed models
A notable release for anyone tracking real-time video generation.
More from Multimodal
- Keeping One Man and His Dog Consistent Across an Entire AI Short Film with Kling 3.0 — Loretaro · 2026-09-16
- MiniMax H3 lands on Together AI: 33B omni-modal model generates 2K clips with native stereo audio — togethercompute · 2026-09-16
- Indie dev replaces flat ghosts with rigged, animated MeshyAI 3D monsters — abandonedmuse · 2026-09-16
- Shareable ChatGPT Image Prompt Keeps Facial Identity While Placing You in a Paris Café Scene — RachelVT42 · 2026-09-16
- FastVideo FastH3 V2 open weights drop with ComfyUI workflow and 8-step version — fruesome · 2026-09-16
- Fully AI-generated GenAI course: ComfyUI lecture made with Index TTS, LTX 2.5 and Codex — CeFurkan · 2026-09-16