Skip to content
ModelVerifiedExternal
Browse files

FLUX.2 Dev FP8 Scaled

flux-dev-non-commercial-license

KindText modelBase modelblack-forest-labs/FLUX.2-devVersionPublisher@silveroxidesBgrade

Model source

Kind
Text model
File size
6.69–30.8 GB
Base model
black-forest-labs/FLUX.2-dev
Version
vRevision a8e566f
Updated
Nov 2, 2026
Publisher
@silveroxides
Downloads
29.4K
Parameters
24B
Tasks
Image edit, Image to image, Text to image
Source
Hugging Face

Source description

You can find workflow with samples in the workflow_assets folder.

The workflow contains the information and links needed to get started with using this model.

This fp8_scaled model is faster than the official one released by ComfyOrg when used with the Loader in the workflow.

The custom node that loads the model in the workflow is necessary to obtain fastest inference on lower VRAM GPU.

If you want to use mxfp8, then you need to use silveroxides/ComfyUI-QuantOps custom node and custom custom comfy-kitchen Repo with instructions until official support is merged.

Use BobJohnson24/ComfyUI-Flux2-INT8 for int8 tensorwise model.

Workflow

Sample

Sources

1 source
Hugging FacePrimaryVersion a8e566fd208b5b6735c1611622fa3999fdfe9b9d
huggingface.co/silveroxides/FLUX.2-dev-fp8_scaled/tree/a8e566fd208b5b6735c1611622fa3999fdfe9b9d

Verified Oct 1

Download files

Model artifacts

24 artifacts

flux-2-klein-4b-dare-merged.safetensors

safetensors · 7.22 GB · SHA-256 fab5aa3d6a40…9342 · Hugging Face

Download file

flux-2-klein-9b-dare-merged-fp8mixedlearnedcomplete.safetensors

safetensors · 8.46 GB · SHA-256 3a76ffdc9efc…c16f · Hugging Face

Download file

flux-2-klein-9b-dare-merged-int8mixedrow_learned.safetensors

safetensors · 8.79 GB · SHA-256 191ad1b8be5b…7828 · Hugging Face

Download file

flux-2-klein-9b-dare-merged-mxfp8mixed_learned.safetensors

safetensors · 9.04 GB · SHA-256 03c3e882739f…db1d · Hugging Face

Download file

flux-2-klein-9b-dare-merged-mxfp8mixed.safetensors

safetensors · 9.04 GB · SHA-256 5bf590ff746b…9f61 · Hugging Face

Download file

flux-2-klein-9b-dare-merged-nvfp4mixedlearned.safetensors

safetensors · 6.69 GB · SHA-256 bb8b36590d24…5ac6 · Hugging Face

Download file

flux-2-klein-9b-dare-merged.safetensors

safetensors · 16.9 GB · SHA-256 c5cfe7952675…5bea · Hugging Face

Download file

flux-2-klein-9b-fp8mixed.safetensors

safetensors · 8.79 GB · SHA-256 55c1ef1fc05e…1850 · Hugging Face

Download file

flux-2-klein-9b-hybrid-mxfp8mixed.safetensors

safetensors · 9.04 GB · SHA-256 06437c2d8cac…4fa6 · Hugging Face

Download file

flux2-dev_fp8matmulmixed.safetensors

safetensors · 30.8 GB · SHA-256 ab9f90ba6f09…cb3b · Hugging Face

Download file

flux2-dev_fp8mixed_NEW.safetensors

safetensors · 30.8 GB · SHA-256 bbac6bd0396c…dd5c · Hugging Face

Download file

flux2-dev-fp8_scaled.safetensors

safetensors · 30.0 GB · SHA-256 4fa310aad3de…7ed5 · Hugging Face

Download file

flux2-dev-fp8mixedfromscaled.safetensors

safetensors · 30.0 GB · SHA-256 641764760ad5…4381 · Hugging Face

Download file

flux2-dev-int8-convrot-simple.safetensors

safetensors · 30.3 GB · SHA-256 7ac2d91aa255…11be · Hugging Face

Download file

flux2-dev-turbo-fp8mixed.safetensors

safetensors · 30.0 GB · SHA-256 cfaafba9d403…7a5a · Hugging Face

Download file

mistral_3_small_flux2_fp8mixed.safetensors

safetensors · 17.2 GB · SHA-256 7b063e0d627d…d66d · Hugging Face

Download file

mistral_small_3_abliterated_v2_fp8mixed.safetensors

safetensors · 16.8 GB · SHA-256 47aea3385314…f626 · Hugging Face

Download file

Mistral-Small-3.2-24B-Instruct-2506-int8.safetensors

safetensors · INT8 (unverified) · 17.1 GB · SHA-256 593222e166f4…211e · Hugging Face

Download file

qwen3_8b_abliterated_v2-fp8mixed.safetensors

safetensors · 7.63 GB · SHA-256 bbcaf05c480c…4ef5 · Hugging Face

Download file

qwen3_8b_abliterated_v2.safetensors

safetensors · 14.1 GB · SHA-256 9d1ef06d5da9…1489 · Hugging Face

Download file

qwen3_8b_abliterated-fp8simplemixed.safetensors

safetensors · 7.68 GB · SHA-256 1b4f088ef6d4…fe54 · Hugging Face

Download file

qwen3_8b_abliterated-int8mixedrow_learned.safetensors

safetensors · 9.08 GB · SHA-256 594963fa9c6c…092b · Hugging Face

Download file

qwen3_8b_abliterated-int8simplemixed-block.safetensors

safetensors · 7.68 GB · SHA-256 f29b22db70e7…81be · Hugging Face

Download file

qwen3_8b_abliterated.safetensors

safetensors · 15.3 GB · SHA-256 52219180c9b1…6b94 · Hugging Face

Download file

Source excerpts

3 excerpts

--- language: - en base_model: - black-forest-labs/FLUX.2-dev license: other license_name: flux-dev-non-commercial-license license_link: LICENSE.md extra_gated_prompt: >- By clicking "Agree", you agree to the [FLUX [dev] Non-Commercial License Agreement](https://huggingface.co/black-forest-labs/FLUX.2-dev/blob/main/LICENSE.txt) and acknowledge the [Acceptable Use Policy](https://bfl.ai/legal/usage-policy). tags: - image-generation - image-editing - flux - diffusion-single-file - comfyui - fp8_scaled pipeline_tag: image-to-image --- ## You can find workflow with samples in the [workflow_assets](https://huggingface.co/silveroxides/FLUX.2-dev-fp8_scaled/tree/main/workflow_assets) folder. ### The [workflow](https://huggingface.co/silveroxides/FLUX.2-dev-fp8_scaled/blob/main/workflow_assets/fp8_scaled_flux2_w_enhanced_prompting-workflow.json) contains the information and links needed to get started with using this model. ### This fp8_scaled model is faster than the official one released by ComfyOrg when used with the Loader in the workflow. ### The custom node that loads the model in the workflow is necessary to obtain fastest inference on lower VRAM GPU. ### If you want to use mxfp8, then you need to use [silveroxides/ComfyUI-QuantOps](https://github.com/silveroxides/ComfyUI-QuantOps) custom node and custom custom comfy-kitchen [Repo with instructions](https://huggingface.co/silveroxides/Chroma1-HD-fp8-scaled/tree/main/experimental) until official support is merged. ### Use [BobJohnson24/ComfyUI-Flux2-INT8](https://github.com/BobJohnson24/ComfyUI-Flux2-INT8) for int8 tensorwise model. ![Workflow](./workflow_assets/fp8_scaled_flux2_w_enhanced_prompting-overview.png) ![Sample](./workflow_assets/fp8_scaled_flux2_w_enhanced_prompting-sample.png)

Source context: 29350 downloads · 95 likes · Pipeline image-to-image · Library diffusion-single-file · Repo silveroxides/FLUX.2-dev-fp8_scaled

Source context: 40971 downloads · 83 likes · Pipeline image-to-image · Library diffusion-single-file · Repo silveroxides/FLUX.2-dev-fp8_scaled