MiniMax open-sources H3, 33B video model with native audio
MiniMax released H3, a 33B-parameter omni-modal model that generates video with native stereo audio at up to 2K resolution and 15-second clips, calling it SOTA open weights. Day 0 support landed in vLLM-Omni, Diffusers, ComfyUI, fal, and LMSYS; community ran it on $280 gaming GPUs and fully-offline MacBooks within 48 hours.
How this story unfolded
12 days · 8 reports · 143 community posts · 151 of 157 shown
- Jul 29
- Jul 31
- Aug 1
- Aug 2
- Aug 3
- Aug 4
- Aug 5
- Aug 7
- Aug 8
- Aug 9
- Aug 10
MiniMax by email
Get an email when MiniMax has news
No news that day, no email.
More stories today
- Discovered Materials raises $9M to find chip-cooling materials with AI
- Your agent didn't hallucinate; it exceeded its authority
- User runs 1M context on 17GB model using 24GB VRAM
- MOSS-TTS-Nano runs realtime multilingual speech on CPU
- The AI Slop Backlash Is Actually Having an Impact