MiniMax H3 Pruned on 16 GB of VRAM
Not at full quality. The 16-bit and 8-bit files are too big for 16 GB; the largest that fits is Q3_K_M (8.9 GB). An official, smaller cut of MiniMax H3 published by Comfy-Org, made for cards that cannot hold the full 33B model.
Q3_K_M (8.9 GB) fits with only 1.3 GB to spare (about 14.7 GB of 16 GB). Quality: noticeable loss of detail. For better quality, Q4_K_M (11.6 GB) also runs, with about 1.4 GB spilling into system RAM - a little slower, still practical.
Also try the FP8 file (21.0 GB) on NVIDIA GPUs. It does not fit entirely, but a current ComfyUI streams the overflow from RAM and native 8-bit files often beat a small GGUF that does fit. Dynamic VRAM →
01Other amounts of VRAM
02Every file, on 16 GB
| File | Size | Needed | On this card | Quality | Download |
|---|---|---|---|---|---|
| 16-bit SAFETENSORS · Comfy-Org | 40.2 GB | 46.0 GB | 30.0 GB too big | the original weights | Hugging Face → |
| Q8_0 GGUF · Abiray | 21.6 GB | 27.4 GB | 11.4 GB too big | practically identical to the original | Hugging Face → |
| FP8 SAFETENSORS · Comfy-Org | 21.0 GB | 26.8 GB | 10.8 GB too big | practically identical to the original | Hugging Face → |
| INT8 SAFETENSORS · Comfy-Org | 21.0 GB | 26.8 GB | 10.8 GB too big | practically identical to the original | Hugging Face → |
| Q6_K GGUF · Abiray | 16.7 GB | 22.5 GB | 6.5 GB too big | very close to the original | Hugging Face → |
| Q5_K_M GGUF · Abiray | 14.1 GB | 19.9 GB | 3.9 GB too big | close; small differences in fine detail | Hugging Face → |
| Q4_K_M GGUF · Abiray | 11.6 GB | 17.4 GB | spills 1.4 GB | good; some loss in fine detail and text | Hugging Face → |
| Q3_K_M ← GGUF · Abiray | 8.9 GB | 14.7 GB | fits · 1.3 GB spare | noticeable loss of detail | Hugging Face → |
“Needed” adds 5 GB of working memory for MiniMax H3 Pruned and a 0.8 GB reserve. Sizes 2026-10-01.
03What to download
| Part | File | Folder | Size | |
|---|---|---|---|---|
| Model | MiniMax-H3-FL2VA-Pruned-Q3_K_M.gguf Q3_K_M · Abiray/MiniMax-H3-Pruned-GGUF | models/unet | 8.9 GB | Download → |
| Text encoder | qwen3vl_32b_minimax_h3_int8_convrot.safetensors Qwen3-VL 32B INT8 (convrot) · Comfy-Org/MiniMax-H3 | models/text_encoders | 27.1 GB | Download → |
| VAE | minimax_h3_video_vae_fp16.safetensors MiniMax H3 video VAE FP16 · Comfy-Org/MiniMax-H3 | models/vae | 5.2 GB | Download → |
| Also needed | minimax_h3_audio_vae_fp32.safetensors MiniMax H3 audio VAE FP32 · Comfy-Org/MiniMax-H3 | models/vae | 0.6 GB | Download → |
| Total download · keep about the same free on disk | 41.9 GB | |||
Alternatives: Qwen3-VL 32B NVFP4 (AWQ) (15.7 GB); Qwen3-VL 32B BF16 (51.5 GB). FP4 files are listed only as alternatives unless the GPU is an RTX 50 (Blackwell). Sizes from Hugging Face, 2026-10-01.
System RAM: 48 GB or more. These files take about 41.9 GB of RAM while loaded, 47.9 GB with Windows and ComfyUI. What happens with less. calculated
04The text encoder
At 15.7 GB even the smallest Qwen3-VL 32B file is too big for 16 GB, so it runs from system RAM. That slows down reading the prompt, not the picture. Sizes: 16-bit 51.5 GB, INT8 27.1 GB, NVFP4 15.7 GB. Why the encoder does not count against VRAM →