Runpod Public Endpoints icon

Runpod Public Endpoints

Call Runpod Public Endpoints (sync/async)

Actions5

Generate Image

AI-generated

Summary

Generate an image using a selected Runpod Public Endpoint model by providing a JSON-formatted input specific to the model's requirements.

Inputs

  • Model (required) — Select the Runpod Public Endpoint model trained or designed for image generation (e.g., Flux, Qwen Image, Seedream).
  • Input JSON (required) — Provide a JSON object with parameters for image generation such as prompt, negative_prompt, width, height, inference steps, guidance, seed, and image_format. This input format varies by model; examples are included in the template and referenced in the Runpod documentation.
  • Download Output as Binary — If enabled, attempts to fetch and attach the generated image binary output by following the output URL; however, currently returns JSON with a download URL.

Output shape

A single JSON object per input item, containing the modelId, the response from the Runpod API, and if requested, a downloadUrl field pointing to the generated image file.

The node supports synchronous and asynchronous calls. If asynchronous, the node can wait and poll for job completion before returning results. Downloading binary data is indicated by a URL in the JSON output; actual binary data fetching is not implemented and must be handled separately.

Examples

Example 1: Generate an image of a serene mountain landscape at sunset.

Select an image generation model like 'Flux Dev', provide an input JSON including "prompt": "A serene mountain landscape at sunset", and optionally enable 'Download Output as Binary' to include the generated image URL in the output.

Links

Discussion