MiniMax H3 on ComfyUI: 10s Video in 9 Mins on Single GPU
sktksm · reddit · 2026-08-03
A developer conducted extensive tests of the MiniMax H3 model in ComfyUI using a single RTX 6000 PRO 96GB, verifying its multimodal reference capabilities. Performance: Generating at 1344x768 took about 3 mins for 2s/5s clips, 9 mins for a 10s clip, and 12.75 mins for a 15s clip. Multimodal Capabilities: Tests included img2video, multi-image reference, and combining 2 images + 3 audio tracks for a 15s clip, maintaining a consistent iteration speed of 9s/it. The author believes this model effectively solves consistency and continuity problems in one shot.
Related event: MiniMax Video Model Goes Viral with Stunning User Tests(61 posts)→
More from Infra
- Trading 2 RTX 5090s for 4 R9700s: does doubled VRAM pay off? — poofph · 2026-10-16
- Running PolyStrata + Cline coding agent on a $0-fancy laptop at ~90 tok/s — rpwoerk · 2026-10-16
- TensorFold 1.0.12 runs 120B vision model on 1-2 DGX Sparks, adds Mac image input — HankYeomans · 2026-10-16
- vLLM brings frontier model local deployment to NVIDIA RTX Spark PCs — vllm_project · 2026-10-16
- NVIDIA signals $2B in new European AI deals at GTC Berlin, totaling ~$10B in a year — CurieuxExplorer · 2026-10-16
- Making Your Data Ready for Agentic AI: trust first, then semantics, then writes — bibryam · 2026-10-16