Floyo FLUX.2 [pro] Custom Node

Custom ComfyUI node for Floyo that wraps the Black Forest Labs FLUX.2 [pro] API. URL-first (no local file paths): you pass URLs and get back a signed output URL for downstream Floyo nodes to download.

Features

• Text-to-image generation with width/height, seed, safety tolerance, and output format controls.
• Image editing with multi-reference support (input_image + up to 7 extra refs).
• Clear URL-only inputs/outputs to match Floyo’s storage pipeline.
• Configurable API key via config.ini or BFL_API_KEY env var.
• Built-in polling with basic error handling and timeouts.

Installation

Repo layout (mirrors Floyo Seed API style):

• nodes/ → node code and helpers
• config.ini → API key placeholder
• requirements.txt
• README.md, __init__.py

1. Place this repository folder (e.g., Flux2-Pro) into ComfyUI/custom_nodes/.
2. Install dependencies:

   cd Flux2-Pro   # or your folder name under custom_nodes
   pip install -r requirements.txt
   

3. Add your key to config.ini or export it:

   [API]
   BFL_API_KEY = <your_real_key>
   BFL_BASE_URL = https://api.bfl.ai/v1/flux-2-pro
   BFL_FLEX_BASE_URL = https://api.bfl.ai/v1/flux-2-flex
   

   export BFL_API_KEY=<your_real_key>
   export BFL_BASE_URL=https://api.bfl.ai/v1/flux-2-pro
   export BFL_FLEX_BASE_URL=https://api.bfl.ai/v1/flux-2-flex
   

4. Restart ComfyUI.

Nodes and I/O

FLUX.2 [pro] Text-to-Image (Flux2ProTextToImage)

• Inputs (required): prompt, width, height
• Inputs (optional): seed (-1 = random), safety_tolerance (0-6), output_format (jpeg/png)
• Output: image (IMAGE tensor ready for Save Image)

FLUX.2 [pro] Image Edit (Flux2ProImageEdit)

• Inputs (required): prompt, input_image (IMAGE tensor)
• Inputs (optional): input_image_2 … input_image_8 (IMAGE tensors, keep total refs <= 8). width, height (0 = match input), seed (-1 = random), safety_tolerance, output_format
• Output: image (IMAGE tensor ready for Save Image)

FLUX.2 [flex] Text-to-Image (Flux2FlexTextToImage)

• Inputs (required): prompt, width, height, guidance (1.5–10), steps (1–50)
• Inputs (optional): seed (-1 = random), safety_tolerance, output_format
• Output: image (IMAGE tensor)

FLUX.2 [flex] Image Edit (Flux2FlexImageEdit)

• Inputs (required): prompt, input_image (IMAGE tensor)
• Inputs (optional): input_image_2 … input_image_10 (IMAGE tensors, keep total refs <= 10), width, height, seed, safety_tolerance, output_format, guidance, steps
• Output: image (IMAGE tensor)

FLUX.2 [pro] Image Edit (Flux2ProImageEdit)

• Inputs (required): prompt, input_image (IMAGE tensor)
• Inputs (optional): input_image_2 … input_image_8 (IMAGE tensors, keep total refs <= 8). width, height (0 = match input), seed (-1 = random), safety_tolerance, output_format
• Output: image (IMAGE tensor ready for Save Image)

Usage Notes

• The returned URL expires quickly; run your Floyo download/output node immediately after this node.
• API statuses handled: Ready, Pending, and error/moderation states. Failures return an error string in the node output.
• Keep total reference images within FLUX.2 [pro] limits (max 8 images, 9MP total). Width/height must be multiples of 16; 0 uses the input image dimensions for edits. Max resolution 4MP (e.g., 2048x2048).
• When connecting ComfyUI image tensors, the node auto-converts to base64. If you prefer URLs, fill the URL fields instead.

API Reference

• Endpoint: https://api.bfl.ai/v1/flux-2-pro
• Docs: https://docs.bfl.ai/quick_start/introduction