Comfy-Org releases MiniMax-H3 weights on HuggingFace
The H3-Base-FL2VA variant accepts zero, one, or two input images, enabling text-to-video, first-frame-to-video, or first-and-last-frame-to-video generation.
1 source
Daily brief
Get tomorrow's AI brief in your inbox
More stories today
- Open-source course teaches phone agent call center with FastRTC and Twilio
- The Rise of the 1 am Job Interview
- Viseron offers self-hosted AI NVR for object and face detection
- Making Knowledge Distillation Cheap Enough to Run at Scale
- How to turn any Claude agent into a 24/7 employee with MCP