MiniMax H3 Pruned NVFP4: Compatible GPUs, VRAM and Download

Main model (diffusion) · NVFP4 · 11.66GB · pruned · updated 2026-09-28

Item	Value
File name	`MiniMax-H3_Ref2VA-NVFP4.safetensors（DmitryDB）`
Component	Main model (diffusion)
Quant / format	NVFP4
Size	11.66 GB
ComfyUI folder	diffusion_models (main model folder)
Source	DmitryDB/MiniMax-H3-ComfyUI-Quants

Which GPUs can run it

736 × 416 · 5s: 46 GPUs workable

GTX 1060 6GB GTX 1070 8GB GTX 1080 8GB GTX 1080 Ti 11GB RTX 2060 6GB RTX 2060 12GB RTX 2060 Super 8GB RTX 2070 Super 8GB RTX 2080 Ti 11GB RTX 2080 Ti 22GB (modded) RTX 3050 8GB RTX 3060 8GB RTX 3060 12GB RTX 3060 Ti 8GB RTX 3070 8GB RTX 3070 16GB (modded) RTX 3070 Ti 8GB RTX 3080 10GB +28 more

1344 × 768 · 5s: 46 GPUs workable

1344 × 768 · 15s: 40 GPUs workable

RTX 2060 12GB RTX 2060 Super 8GB RTX 2070 Super 8GB RTX 2080 Ti 11GB RTX 2080 Ti 22GB (modded) RTX 3060 8GB RTX 3060 12GB RTX 3060 Ti 8GB RTX 3070 8GB RTX 3070 16GB (modded) RTX 3070 Ti 8GB RTX 3080 10GB RTX 3080 12GB RTX 3080 20GB (modded) RTX 3080 Ti 12GB RTX 3090 24GB RTX 4060 8GB RTX 4060 Ti 8GB +22 more

“Workable” = estimated status ok or warn with 32GB RAM / fast-disk (warn means weight offloading and noticeably slower).

Every quant in this family, side by side

Same family = different quants of the same base variant. They are interchangeable — what changes is size, speed and quality loss. "Peak" is a rough VRAM estimate (file size × 0.6, floor 8G).

File	Quant	Size	Pruned	VRAM fit
P- NVFP4 ← this page	NVFP4	11.66 GB (peak ~8G)	pruned	for 12-16GB cards
P- W4A8	W4A8	11.77 GB (peak ~8G)	pruned	for 12-16GB cards
P- FP8	FP8 Scaled	20.96 GB (peak ~13G)	pruned	wants 24GB+
P- INT8	INT8 ConvRot	20.97 GB (peak ~13G)	pruned	wants 24GB+
INT8	INT8 ConvRot	34.04 GB (peak ~20G)	full	wants 24GB+
P- BF16	BF16	40.23 GB (peak ~24G)	pruned	wants 24GB+
BF16	BF16	66.28 GB (peak ~40G)	full	wants 24GB+

Rule of thumb: anything under 10GB fits an 8GB card, the 10-16GB tier targets 12-16GB cards, and files above 16GB want 24GB+. Within this family the flagged P- INT8 is the official default — take it unless you have a reason not to. The smallest here is P- NVFP4 (11.66GB) and the largest is BF16 (66.28GB); this file (11.66GB) sits in the "for 12-16GB cards" tier. About this quant: NVFP4: 4-bit float — the main model needs an RTX 50 (Blackwell) card. "Pruned" trims redundant channels for size — the quality cost is usually tiny.

Who is this file for?

Role in the pipeline: the main model — it sets generation quality and speed and is the biggest download; swapping quants changes detail and pace, not how you write prompts.Aimed at 12-16GB cards (4070 / 4070 Ti Super / 5060 Ti 16G class): no offloading, the best speed-for-quality balance. On 8GB? Look at the under-10GB entries in the table instead.

Fastest vs. minimum hardware

"Fastest" = the card with the shortest estimated time; "floor" = the smallest-VRAM card that still runs it. Times include model load and text encoding; the first run also compiles kernels.

736 × 416 · 5s: fastest RTX 5090 32GB at ~47 s (46/49 GPUs workable) · floor GTX 1060 6GB (running right at the VRAM ceiling)
1344 × 768 · 5s: fastest RTX 5090 32GB at ~1 min 56 s (46/49 GPUs workable) · floor GTX 1060 6GB (running right at the VRAM ceiling)
1344 × 768 · 15s: fastest RTX 5090 32GB at ~10 min 30 s (40/49 GPUs workable) · floor RTX 3060 8GB (running right at the VRAM ceiling)

Download, place, refresh — 3 steps

Grab it: the button above redirects to the official HuggingFace / GitHub repo — /go is just a redirect, we never rehost. Skip cloud-drive mirrors; if HuggingFace is slow or blocked on your network, a mirror can speed up the transfer, but verify the file against the official repo. Keep at least 2× the file size free on disk.
Place it: drop the file into diffusion_models (main model folder) exactly as shown in the spec table, filename untouched — loaders locate files by folder and name, so a rename makes it "vanish".
Load it: hit Refresh in ComfyUI (it only scans folders at startup or on refresh), then pick the file in the loader node. loader node: Load Diffusion Model (GGUF files go through the ComfyUI-GGUF Unet Loader).

Sanity check: once downloaded, compare the size on disk with the 11.66GB listed here. Way off? The download is incomplete — browsers sometimes keep a half-written file looking like a finished one — so delete it and pull again.

Also required: It takes all three to render: besides this file you need the text encoder NVFP4 AWQ and the VAEs BF16 + BF16 — miss any one of them and the workflow errors out on the spot.

Won't load? Check these three

Nothing shows up after Refresh: nine times out of ten it's the wrong folder — re-check the ComfyUI folder row in the spec table above. The file must sit inside ComfyUI's models/ subfolder, and no, a desktop shortcut doesn't count.
Instant OOM the moment it loads: make sure no other big model is sitting in VRAM first, then step down to a smaller quant from the table above — the INT8 vs BF16 guide has the full math on why brute-forcing backfires.
Renamed it and now nothing picks it up: loaders match by filename (GGUF metadata stores the original name too), so restore the original name and hit Refresh once more.

Download · HuggingFace →

→ Run the numbers for your RAM/budget in the calculator (this file pre-filled)

FAQ

How much VRAM does MiniMax-H3_Ref2VA-NVFP4.safetensors（DmitryDB） need?

The file is 11.66GB (Main model (diffusion)). Runtime VRAM also depends on resolution and duration; a 720p-class short clip typically needs roughly 8GB+ of free VRAM. Run the calculator for your exact card.

Where does this file go in ComfyUI?

diffusion_models (main model folder). Keep the filename as-is; refresh ComfyUI and select it in the matching loader.

How does it compare with other quantizations?

See sibling entries in the model library and the INT8/BF16 comparison article. Rule of thumb: 8GB → GGUF Q3/Q2; 12-16GB → pruned INT8 or W4A8; 24GB+ → full INT8 or NVFP4 (RTX 50).

Similar files: BF16 · INT8 · P- BF16 · P- INT8 · P- FP8 · BF16 · INT8 · P- BF16 · P- INT8 · P- FP8