DEVELOPER REFERENCE · VERIFIED MODEL EVIDENCE
Comfy-Org
Compare 24 source-verified model records, repository licenses, published weight formats and dated evidence. No overall model ranking.
Evidence statusSource-first verified
Latest record check2026-11-13
Latest record full verification2026-11-03
DIRECT ANSWER · SOURCE-OBSERVED REGISTRY
What Comfy-Org model evidence is available?
Open Model Weights currently tracks 24 field-verified Comfy-Org model records. In this indexed sample, the recorded modalities are 23 Not declared; 1 Image To Video. The published repository license labels include 12 apache-2.0; 9 other; 2 mit; 1 creativeml-openrail-m. The listed weight formats include Safetensors (24).
Coverage describes the model repositories indexed here, not every model released by the publisher. A checked source, an unknown field and an unchanged source revision are different states. No missing value is treated as a negative answer.
24verified model records
2recorded modalities
4distinct license labels
24model repository links
Comfy-Org maintains model-file repositories prepared for ComfyUI workflows. Its documented releases frequently repackage checkpoints originally published by other model developers, including Wan, Qwen, FLUX and video or audio model projects.
A Comfy-Org namespace denotes the distributor or curator of a particular artifact repository, not necessarily the original model creator. Modified packing, quantization, supplementary encoders, LoRA files and workflow templates are different claims from authorship or original model training.
01Why Comfy-Org repositories are not a model family
Comfy-Org publishes collections of model files for the ComfyUI graph-based interface. Its Wan 2.2 repackaging model card points to many upstream repositories and provides exact ComfyUI file paths. The upstream developer must be retained for provenance; the Comfy-Org record represents the verified distributable artifacts.
02ComfyUI directory layout and dependency files
Some workflows need separate diffusion checkpoints, text encoders, VAE files, audio encoders or LoRA adapters. The publisher's Wan 2.2 file tree explicitly lists these categories and provides sample workflow links. Downloading just one .safetensors artifact is not always enough to reproduce a working pipeline.
03Image, video and audio packaging differ
Tracked repositories include FLUX image generation, Qwen-Image material, Wan video workflows, and audio or animation assets. For many packages the current registry modality is not declared; the repository name alone should not be promoted into a verified capability. Follow the source card and reference ComfyUI workflow before configuring a model.
04License labels and inherited restrictions
A repackaging may carry the original model's license or additional distribution obligations. In this registry Comfy-Org packages include Apache-2.0, custom, OpenRAIL and other labels. A permissive label on one helper artifact cannot establish a commercial right to every upstream checkpoint in the workflow; inspect each original source and model card.
05ComfyUI compatibility and trust boundaries
The ComfyUI documentation describes workflow nodes and model setup. A repository built for this interface is not necessarily a runnable model on vLLM or llama.cpp, nor evidence of independent runtime benchmarks. Pin the file hashes and installed custom-node versions and treat missing compatibility evidence as unknown.
Publisher background is attributed to official sources. Model counts and license classifications come from OMW repository evidence; no model benchmark or processing location is inferred.
Image To Videoapache-2.0
Wan 2.2 video workflow files
Inspect the upstream model list, diffusion files, text encoders and workflow templates before downloading.
Not declaredother
Qwen Image artifact packaging
Review this Comfy-Org Qwen-Image release separately from the original Qwen publisher record.
Not declaredother
FLUX.1 development weights
Check the original FLUX [dev] licensing and the ComfyUI-specific packaging of the weights.
Not declaredapache-2.0
FLUX.1 schnell distribution
Check the separate FLUX.1-schnell record and its Apache-2.0 source classification.
Not declaredother
FP8 image model package
Inspect the Stability AI community license and inference workflow; FP8 is a format signal, not a speed measurement.
Not declaredapache-2.0
Audio and music workflow files
Use the ACE-Step 1.5 repackaging card for correct model folder placement and original publisher attribution.
Compare models →Explore weight-memory estimates →Understand licensing →
FIELD-BY-FIELD COMPARISON · PRIMARY LINKS
Model evidence at a glance.
Compare the exact entries in our registry. Parameter values are taken from the source-checked record; a license label is not a legal opinion, and recognized weight files are not an inference-speed benchmark. Open each record for the underlying citations and observed history.
Compare two to four models →Explore weight memory →Browse observed history →
LICENSE & COVERAGE · NOT LEGAL ADVICE
How do the recorded licenses differ?
Different models from the same developer can use different license terms. The distribution below describes the labels and field-level classifications observed in the listed repositories; it does not automatically establish permissions for a specific deployment.
License labels
- apache-2.012 of 24
- other9 of 24
- mit2 of 24
- creativeml-openrail-m1 of 24
Commercial-use classifications
- allowed14 of 24
- unknown10 of 24
Model modalities
- Not declared23 of 24
- Image To Video1 of 24
EVIDENCE METHOD · TRANSPARENT LIMITS
How this developer profile is assembled.
We aggregate the current source-checked model records, rather than publish an unverified corporate biography. Repository metadata, model cards, license files and recognized weight artifacts supply the facts where available. The record-check date does not assert that a model or its license changed on that date.
Source-derived runtime signals are not hardware benchmarks. Weight-only memory estimates omit runtime overhead, activations and other deployment requirements. Model-specific uncertainties remain visible in the individual evidence records.
Evidence policy →Full methodology →Independent measurements →
QUESTIONS · DIRECT ANSWERS
Frequently asked questions about Comfy-Org models.
Which Comfy-Org models are covered?
Open Model Weights currently lists 24 field-verified records for Comfy-Org. Their recorded model modalities are: 23 Not declared; 1 Image To Video. This is a registry sample, not an exhaustive publisher catalog.
Which licenses are recorded for Comfy-Org models?
In the 24 indexed records, repository license labels are: 12 apache-2.0; 9 other; 2 mit; 1 creativeml-openrail-m. License labels do not replace the model-specific terms or establish rights for every use case.
Where can I verify Comfy-Org model information?
Each indexed model links to its official source repository and a field-level Open Model Weights profile. The most recent record checks shown here reach 2026-11-13; checking a repository does not imply that its contents changed.
Does Comfy-Org develop every model it publishes?
No. Many Comfy-Org model cards explicitly state that checkpoint files were repackaged and link to original developers. Preserve both the upstream author and the distribution repository as separate evidence.
Can every Comfy-Org .safetensors file run by itself?
No. A ComfyUI workflow can require several weight files, text encoders, a VAE, audio encoders or adapters. The official Wan 2.2 repackaging card lists distinct directories and dependencies.
Are all Comfy-Org repackaged weights free for commercial use?
No blanket permission exists. Each upstream model and packaged artifact can have different licensing terms. Our registry keeps custom and unresolved labels visible instead of classifying them as automatically allowed.
COMPLETE INDEXED SELECTION
All 24 indexed model profiles.