We have added the sync parameter to support synchronous return of results for easier debugging. When set to true, the result will be returned synchronously.
The tolerance level for content review of input and output. A higher level means looser review restrictions.
webhook_url
The URL to receive Webhook notifications for asynchronously obtaining generation results.
webhook_secret
An optional secret key for Webhook signature verification to enhance request security.
steps
The number of iteration steps for image generation. More steps result in richer image details but slower generation speed.
prompt_upsampling
Whether to perform upsampling optimization on the prompt to improve prompt understanding accuracy.
guidance
Guidance strength, which controls the degree of alignment between the generated result and the prompt. A higher value means higher alignment but lower creativity.
interval
The interval for outputting intermediate results during generation, used to check the generation progress.
📝 Text-driven generation: Generate high-quality images based on prompts, suitable for various creative scenarios such as artistic creation, product design, and scene rendering.
🖼️ Multi-image assisted creation: Supports up to 8 input images for assisted generation, enabling advanced effects like image fusion, style transfer, and content expansion.
⚙️ Highly customizable configuration: Allows customization of image height, output format, and flexible control of content review strictness through the safety tolerance parameter.
📊 Asynchronous task management: Returns task results asynchronously, including key information such as task ID and polling address, facilitating task tracking.
Steps for Using the Asynchronous Interface:
1.
Use the generation interface (/flux/v1/flux-2-max), fill in the corresponding parameters, and obtain the task ID.
2.
Use the query interface (/flux/v1/get_result) and fill in the task ID to obtain the result.