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: Reducto
providerId: reducto-ai
created: '2026-05-25'
modified: '2026-05-25'
reconciled: true
tags:
  - Rate Limiting
  - Quotas
  - Document AI
description: 'Reconciled rate limits for the Reducto API. Sources: docs.reducto.ai/reference/rate-limits and
  reducto.ai/pricing. Sync endpoints share a 200-concurrent-request pool; async endpoints accept submissions
  up to 500 RPS and bypass concurrency by returning job IDs.'
sources:
  - https://docs.reducto.ai/reference/rate-limits
  - https://reducto.ai/pricing
responseCodes:
  throttled: 429
  quotaExceeded: 429
algorithm: concurrency-and-submission-rate
limits:
  - tier: Standard
    syncConcurrent: 200
    submissionRPS: 1
    notes: Pay-as-you-go plan. 1 sync request per second.
  - tier: Growth
    syncConcurrent: 200
    submissionRPS: 10
    priorityRequests: 5
    notes: Premium rate limit. Up to 5 active priority requests.
  - tier: Enterprise
    syncConcurrent: 200
    submissionRPS: 100
    notes: 100+ RPS with custom throughput available on request.
endpoints:
  sync:
    - /parse
    - /extract
    - /split
    - /edit
    - /pipeline
    - /classify
    - /cite
    - /upload
  async:
    - /parse_async
    - /extract_async
    - /split_async
    - /edit_async
    - /pipeline_async
  notes: Async endpoints return immediately with a job_id, bypassing the sync concurrency cap. Submission
    rate is still bound by the per-tier RPS limit.
retryGuidance: Official Python and Node SDKs implement exponential backoff on 429 automatically.