MiniMax H3 open weights released, SOTA open video model

How this story unfolded

4 weeks · 5 reports · 111 community posts · 116 of 123 shown

1. Jul 31
   • 9 community posts
2. Aug 1
   • MiniMax Releases MiniMax H3: An Omni-Modal Video Model That Generates 15-Second 2K Clips With Native Stereo Audiomarktechpost.com
   • 5 community posts
3. Aug 2
   • 3 community posts
4. Aug 3
   • MiniMaxAI/MiniMax-H3 · Hugging Facehuggingface.co
   • MiniMax H3 - a multimodalart Collectionhuggingface.co
   • Comfy-Org/MiniMax-H3huggingface.co
   • 11 community posts
5. Aug 4
   • 4 community posts
6. Aug 5
   • 3 community posts
7. Aug 6
   • ComfyUI MiniMax H3: Best Video Generation Workflows (Ep29)
8. Aug 7
   • 2 community posts
9. Aug 10
   • MiniMax-H3 local benchmark: six optimization stacks on the INT8 build, 28 prompts, 168 clips - and an open blind vote
10. Aug 11
    • Implementing a MiniMax-H3 Multimodal Video and Audio Generation Pipeline with ComfyUI APIsmarktechpost.com
    • 2 community posts
11. Aug 13
    • 4 community posts
12. Aug 14
    • 4 community posts
13. Aug 21
    • The H3 dialog prompting guide sucks
14. Aug 22
    • H3 can do Side-by-Side VR/3D Videos natively
15. Aug 23
    • Sparse Attention, Harder, Better, Faster, Stronger
16. Aug 24
    • 3 community posts
17. Aug 25
    • 6 community posts
18. Aug 26
    • 12 community posts
19. Aug 27
    • 10 community posts
20. Aug 28
    • 15 community posts
21. Aug 29
    • 10 community posts
22. Aug 30
    • 3 community posts