MiniMax releases MiniMax-H3, SOTA open-weights video model
MiniMax-H3 is now the #1 open video generation model on Video Arena for both text-to-video and image-to-video, tied #3 overall. Within 48 hours the community ran it on RTX 5090s and offline MacBooks, with ComfyUI nodes live at launch. Deployment in the US, EU, UK, and South Korea requires formal authorization beyond the community license.
15 sources
ethanfel/Qwen3-VL-32B-Ultra-Heretic-H3-ComfyUI-INT8-ConvRothuggingface.co
We just dropped SOTA open weights for our MiniMax‑H3 video model, and so much happened in 48 hours....x.com
PipeNetwork/minimax-h3-mlxsimonwillison.net
The MiniMax-H3 video generation model is a lot of fun - here's what I generated on my M5 Pro Mac for...bsky.app
MiniMax Releases MiniMax H3: An Omni-Modal Video Model That Generates 15-Second 2K Clips With Native Stereo Audiomarktechpost.com
Minimax is out so you know what this meansreddit.com
MiniMax by email
Get an email when MiniMax has news
No news that day, no email.
More stories today
- Shopify reports AI search tripled traffic and sales in Q2
- Hark unveils Handoff, a browser use agent for everyday web tasks
- Brookfield Asset Management reports record fundraising driven by AI demand
- Complete AI Agent Building Blueprint details agent development
- TechCrunch Disrupt 2026 announces Real World AI Stage