Heads up: statuses are automated heuristics from commit dates, registry metadata and issue titles. "Stale" can mean "finished and stable"; "Error spike" means error-titled issues are unusually high for that pack lately — worth checking recent issues, not a verdict that it is broken. Always check the repo.

ComfyUI-GGUF Active

GGUF Quantization support for native ComfyUI models.

Why: last commit 280 days ago; Quiet: no commits in 6+ months

Repo	city96/ComfyUI-GGUF (default branch main)
Registry	ComfyUI-GGUF · publisher city96 · latest 1.1.10 (2026-01-12) · in Manager list: yes
Popularity	rank #9 · 2,700,438 downloads · 4155 stars
Tags	Loaders, quantization & speed Generic / any
Star velocity	collecting (nightly snapshots started 2026-10-19)
Error issues	1 error-titled issues opened in last 30 days (open+closed) · 0.04 per 100K downloads · baseline 0.83/month over prior 6 months (5 total) · 1.0× baseline (Error spike needs ≥3, ≥2× baseline, ≥0.5 per 100K)
Last commit	2026-01-12 (280 days ago)
Open issues	66+ issues, 34 PRs (fetched 2026-10-19)
pyproject.toml	present · version 2.0.0 · requires-comfyui — · frontend — · requires-python — · 3 deps
Registry ComfyUI req.

Active forks

Automatic suggestions, not endorsements: forks with commits newer than upstream, ranked by commits ahead, recency and stars. Read the fork's README and diff before switching. Scanned 2026-10-19: 166 forks checked.

Fork	Commits	Last commit	Stars
molbal/ComfyUI-GGUF Merge pull request #33 from molbal/feature/add-model-moss	100 ahead / 0 behind	2026-10-18	95 ★
ChrisColeTech/ComfyUI-GGUF-Loader 2.20.0: TRELLIS.2 + Pixal3D from the owner's per-stage files (safetensors + GGUF) into core's native pipeline	166 ahead / 0 behind	2026-10-15	24 ★
Nif00/ComfyUI-GGUF Fix MiniMax H3 GGUF loading	46 ahead / 0 behind	2026-08-08	10 ★

Error-titled issues, last 30 days

Qwen-Image 2.1 fails with GGUF: "mat1 and mat2 shapes cannot be multiplied (215x24576 and 12288x4096)" (`input_act` / `residual` ignored by GGMLOps.Linear) 2026-10-07 · open

Open issues mentioning errors / breakage

#487 Qwen-Image 2.1 fails with GGUF: "mat1 and mat2 shapes cannot be multiplied (215x24576 and 12288x4096)" (`input_act` / `residual` ignored by GGMLOps.Linear) 2026-10-07
#475 Failed in installing ComfyUI-GGUF nightly version 2026-08-16
#457 np.uint32(41) is not a valid GGMLQuantizationType error on using bonsai_image_4b-q1_0.gguf 2026-06-09
#453 GGUF broken after Comfyui latest update 2026-06-07
#444 [Bug Report] Windows fatal exception: access violation when node has multiple outputs 2026-05-08
#443 ERROR: Could not open requirements file: [Errno 2] No such file or directory: '.\\ComfyUI\\custom_nodes\\ComfyUI-GGUF\\requirements.txt' 2026-05-08
#431 z_image_base_q8_0.gguf gives Runtime Error 2026-03-17
#430 RuntimeError: CUDACachingAllocator INTERNAL ASSERT FAILED on Jetson Orin NX (CUDA 12.6, PyTorch nv24.08/2.8.0) 2026-03-15
#424 Doesn't work with new "TextGenerate" node (add mmproj support) 2026-03-01
#421 Stability and running speed have been improved. Unfortunately, it is not possible to use AI to convert comments, as this would cause errors. 2026-02-22
#420 Fixed the stability issue that caused crashes when switching models multiple times during loading. 2026-02-21
#417 Error(s) in loading state_dict for NextDiT 2026-02-07
#401 comfy ui error 2026-01-08
#379 GGUF size mismatch error on Mac 2025-12-08
#354 CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect. 2025-11-01

Recent commits

2026-01-12 [main] Only include metadata on new comfy versions
2026-01-12 [main] Add Gemma3 12B Support (#402)
2026-01-12 [main] update loader.py
2026-01-11 [main] Update loader.py
2026-01-11 [main] Support tokenizer recreation from metadata