Run an inline prompt (async, no pre-created job)

One-call run: pass a prompt (with optional {{placeholders}} filled by inputs) and an optional output_schema. Returns 202 immediately; poll GET /runs/{id} or supply callback_url. project_id is optional — omit it and the run executes in the org's auto-created "API Runs" project.

post/runs

Request

  • Base URL: https://worker.openhelm.ai/v1
  • URL: https://worker.openhelm.ai/v1/runs
  • Auth: HTTP bearer

Request body

promptstring required
inputsobject
output_schemaobject

Declares the result shape.

model'haiku' | 'sonnet' | 'opus'
project_idstring

Optional; defaults to the org's "API Runs" project.

namestring

Optional label for the run.

callback_urlstring uri

HMAC-signed completion webhook target.

idempotency_keystring

Replays return the original run.

Example request

{
  "prompt": "Research {{company}} and summarise its latest funding round, investors, and valuation."
}

Response

Accepted

task_idstring
status'queued'
poll_urlstring

Changes

No recorded changes to this operation across all 1 revision of this API.