RTX 5090 Generates 15s Videos in 2 Minutes: Minimax H3 Speed-Tuned ComfyUI Settings Shared
Free_Pressure8623 · reddit · 2026-11-02
A Reddit user shares ComfyUI settings balancing speed and quality for MiniMax H3: OliviaRossi's int8 quantized model plus the Veda Sparse Attention node (generalizedsparsity and referencesparsity both at 90%), 6 steps, icm sampler, beta57 scheduler at 0.6MP followed by 1.5x upscale. On an RTX 5090 with 64GB RAM this yields 15-second videos in 2 minutes while keeping audio quality high. Author notes it's tuned for speed, not maximum realism.
More from Multimodal
- Tencent ships three speech models: ASR, TTS and real-time voice-preserving translation — solyarisoftware · 2026-11-02
- Fibo 2.0 image editing ships a 2-step distilled version on Hugging Face — Illustrious_Row_9971 · 2026-11-02
- 4B Qwen3.5 fine-tune generates game-ready 3D assets with one-click GLB export — linoy_tsaban · 2026-11-02
- Krea2 character training: swapping LoRA for LoKR tanks results, author seeks fix — AkaToraX · 2026-11-02
- Suno Studio adds track hiding, grid controls, and fine pitch tuning — suno · 2026-11-02
- Suno Studio update adds clip settings sidebar, fine pitch control and auto-stretch options — suno · 2026-11-02