generated: '2026-07-20'
method: searched
source: https://docs.priorlabs.ai/api-reference/getting-started
authentication:
  style: bearer
  header: Authorization
  format: Bearer <api_key>  # JWT
  docs: https://docs.priorlabs.ai/api-reference/getting-started
  ref: authentication/priorlabs-authentication.yml
idempotency:
  supported: false
  note: No idempotency-key header is documented; fit/predict are stateful jobs keyed by uploaded dataset ids.
pagination:
  supported: false
  note: Endpoints are RPC-style (fit/predict/upload); no list/collection pagination surface.
data_transfer:
  pattern: signed-url-upload
  note: prepare_train_set_upload / prepare_test_set_upload return signed cloud-storage URLs; clients PUT CSV/Parquet, then call fit/predict by dataset id.
  formats: [csv, parquet]
error_envelope:
  schema: ErrorResponse
  fields: [code, detail, retryable, support]
  ref: errors/priorlabs-problem-types.yml
rate_limiting:
  signal: HTTP 429
  model: daily + monthly token pools; separate monthly thinking-fit quota
  daily_token_limit: 50000000
  monthly_token_limit: 200000000
  monthly_thinking_fit_quota: 20
  docs: https://docs.priorlabs.ai/api-reference/metering
versioning:
  ref: lifecycle/priorlabs-lifecycle.yml