specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Extend
providerId: extend-ai
created: '2026-06-20'
modified: '2026-06-20'
reconciled: false
tags:
- Document Processing
- Document AI
- Intelligent Document Processing
- OCR
- Rate Limiting
- Quotas
- Throttling
description: >-
  The Extend API enforces account- and workspace-scoped throughput controls.
  Synchronous processor endpoints (/parse, /extract, /classify, /split) carry a
  five minute per-request timeout and are intended for onboarding and testing,
  while production traffic is expected to use the asynchronous run endpoints,
  batch endpoints, and webhooks. Workflow runs accept a priority value (default
  50) that influences scheduling under load. Specific numeric request- and
  credit-per-minute limits are not publicly enumerated and are managed per
  account and plan, so this artifact is not reconciled.
notes: >-
  Verify concurrency, requests-per-minute, and batch sizing limits with Extend
  on reconciliation. Synchronous endpoints time out at five minutes; use async
  runs, batches, and webhooks for production scale.
sources:
- https://docs.extend.ai/api-quickstart
- https://docs.extend.ai/api-reference/endpoints/parse/parse-file
- https://docs.extend.ai/webhooks/configuration
responseCodes:
  throttled: 429
limits:
- name: Synchronous Request Timeout
  scope: request
  metric: seconds
  limit: 300
  notes: Sync /parse, /extract, /classify, /split fail after five minutes; use async runs.
- name: Requests Per Minute
  scope: account
  metric: requests
  limit: see provider documentation
  notes: Account- and workspace-scoped; not publicly enumerated.
- name: Concurrent Runs
  scope: account
  metric: runs
  limit: see provider documentation
  notes: Concurrency governed per account and plan; use batch endpoints for high volume.
- name: Workflow Run Priority
  scope: account
  metric: priority
  limit: 50
  notes: Default priority value on workflow runs; higher priority is scheduled sooner under load.
- name: Run Metadata Size
  scope: request
  metric: bytes
  limit: 10240
  notes: Custom run metadata is capped at roughly 10KB per run.