Runpod Public Endpoints icon

Runpod Public Endpoints

Call Runpod Public Endpoints (sync/async)

Actions5

Generate Text

AI-generated

Summary

Generate text synchronously or asynchronously using a selected Runpod Public Endpoint model and a JSON input formatted according to the model's specifications.

Inputs

  • Model (required) — Select a Runpod Public Endpoint text generation model from a dynamic list that updates based on available models.
  • Input JSON (required) — Provide a JSON string specifying the input parameters formatted according to the selected model's requirements, e.g., messages and sampling parameters for chat models.
  • Download Output as Binary — Optionally fetch output files (e.g., images or audio) linked by URL and attach the content as binary data in the workflow output.

Output shape

a list of JSON objects, each containing the API response merged with the model ID and optionally a downloadUrl property if download is enabled.

If download is enabled, the node adds a downloadUrl field with the URL of image, audio, or file output. Actual binary download is noted as TODO. Supports processing multiple input items individually.

Examples

Example 1: Generate text response using the Granite or Qwen3 chat model.

Set operation to 'Generate Text', choose a text generation model, and provide the input JSON with conversation messages and sampling parameters based on the Runpod documentation.

Links

Discussion