Technical Specifications of flux-pro-1-0-canny
| Specification | Details |
|---|---|
| Model ID | flux-pro-1-0-canny |
| Model family | FLUX.1 Canny [pro] |
| Provider / origin | Black Forest Labs’ FLUX family, also exposed through third-party inference platforms such as Replicate and fal. |
| Modality | Image-to-image / controlled image generation using a text prompt plus a Canny edge control image. |
| Core control method | Canny edge detection for structure-preserving guidance. |
| Primary use cases | Sketch-to-image, retexturing while preserving composition, controlled style transfer, and architectural or layout-constrained visualization. |
| Input types | Prompt text plus a control image / edge map; common runtime parameters on public deployments include steps, guidance, seed, output format, and safety-related controls. |
| Output | Generated image files following the structure of the supplied edge map. |
| Inference pattern | Typically asynchronous task submission with result polling on FLUX-style APIs. |
What is flux-pro-1-0-canny?
flux-pro-1-0-canny is CometAPI’s platform identifier for the FLUX.1 Canny [pro] model, a professionally oriented edge-guided image generation model in the FLUX ecosystem. Its defining capability is that it uses Canny edge detection as a structural conditioning signal, so the model can generate a new image that follows the composition and contours of an input image while still responding to a natural-language prompt.
In practical terms, this model is designed for users who want more control than standard text-to-image generation provides. Instead of describing everything purely in words, you can provide a sketch, extracted edge map, or composition reference and ask the model to render it in a different style, material treatment, or scene context. Public model descriptions consistently position it as especially useful for preserving layout and structure during transformation tasks.
Because the underlying FLUX APIs often use job submission plus polling, integrations with this kind of model are commonly built as asynchronous workflows: submit the request, receive a task ID, then fetch the result once processing is complete.
Main features of flux-pro-1-0-canny
- Edge-guided structural control: The model uses Canny edge maps to preserve the composition, outlines, and major forms of the source image during generation.
- Prompt-driven restyling: You can combine an edge-controlled input with a text prompt to change appearance, mood, materials, or artistic style without discarding the original structure.
- Strong fit for sketch-to-image workflows: Public examples and descriptions highlight turning sketches or edge-derived guides into polished artwork and production-ready visuals.
- Composition-preserving retexturing: It is well suited for retexturing scenes or objects while keeping the layout stable, which is useful in product, concept, and environment visualization workflows.
- Useful parameter control: Public API surfaces for this model commonly expose inference settings such as diffusion steps, guidance strength, seed, output format, and safety tolerance, enabling reproducibility and tuning.
- Production-friendly API behavior: FLUX API integration guidance emphasizes endpoint-based job submission and result polling, which fits backend automation and application pipelines well.
How to access and integrate flux-pro-1-0-canny
Step 1: Sign Up for API Key
Sign up on CometAPI and generate your API key from the dashboard. Store it securely and load it through an environment variable in your application before making requests to flux-pro-1-0-canny API.
Step 2: Send Requests to flux-pro-1-0-canny API
Use CometAPI's Flux-compatible endpoint at POST /flux/v1/{model}.
curl https://api.cometapi.com/flux/v1/flux-pro-1-0-canny \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $COMETAPI_API_KEY" \
-d '{
"prompt": "A breathtaking landscape painting with golden hour lighting",
"width": 1024,
"height": 768,
"steps": 20,
"prompt_upsampling": true
}'
Step 3: Retrieve and Verify Results
The API returns a task ID. Poll GET /flux/v1/get_result?id={task_id} to check generation status and retrieve the output image when complete.