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: Sentinel Hub
providerId: sentinel-hub
created: '2026-06-13'
modified: '2026-06-13'
reconciled: true
tags:
  - Satellite Imagery
  - Geospatial
  - Rate Limiting
  - Processing Units
description: >-
  Sentinel Hub enforces per-minute rate limits on both Processing Units (PUs) and HTTP requests.
  All limits are minute-based and must simultaneously be satisfied. Unused PUs and requests
  do not accumulate to the next minute or month. Exceeding either limit returns HTTP 429 with a
  Retry-After header (milliseconds until next available slot). Exceeding monthly quota returns
  HTTP 403. Trial accounts are limited to 10 requests and 10 PUs per minute.
notes: >-
  Subscription tiers determine the exact PU/minute and requests/minute caps. Contact Planet/
  Sentinel Hub support to request quota adjustments. Spread requests evenly across the minute
  (e.g. one every 100ms for a 10 RPM plan) to maximise throughput.
sources:
  - https://docs.sentinel-hub.com/api/latest/api/overview/rate-limiting/
  - https://docs.sentinel-hub.com/api/latest/api/overview/billing/
  - https://documentation.dataspace.copernicus.eu/APIs/SentinelHub/Overview/ProcessingUnit.html
responseCodes:
  throttled: 429
  quotaExceeded: 403
  headers:
    retryAfter:
      name: Retry-After
      description: Time in milliseconds until the next request slot is available.
limits:
  - name: Processing Units per Minute
    scope: account
    metric: processing-units-per-minute
    limit: varies by subscription tier (10 for trial)
    timeFrame: minute
    notes: >-
      All subscription types use minute-based PU limiting. Batch Processing API requests
      consume 1/3 PUs; Async Processing API requests consume 2/3 PUs for areas >=10,000 px.
  - name: HTTP Requests per Minute
    scope: account
    metric: requests-per-minute
    limit: varies by subscription tier (10 for trial)
    timeFrame: minute
    notes: >-
      Separate from PU limit; both must be satisfied simultaneously. Token (OAuth2) requests
      themselves are also rate-limited — reuse tokens within their validity window.
  - name: Processing Units per Month
    scope: account
    metric: processing-units-per-month
    limit: varies by subscription package
    timeFrame: month
    notes: >-
      Monthly allocation resets on the 1st of each calendar month regardless of subscription
      start date. Unused PUs are forfeited at reset.
  - name: OAuth2 Token Requests
    scope: account
    metric: token-requests-per-minute
    limit: rate-limited (exact value not published)
    timeFrame: minute
    notes: >-
      Excessive token requests trigger HTTP 429. Reuse JWT access tokens for their full
      validity period rather than fetching a new token per API call.
policies:
  - name: Burst Spreading
    description: >-
      Distribute requests evenly across the rate-limit window rather than bursting at the
      start of the minute. For a 10 RPM plan, send one request every 6 seconds.
  - name: Exponential Backoff
    description: >-
      On HTTP 429, wait for the Retry-After interval then retry with exponentially increasing
      back-off and random jitter to avoid thundering-herd after a rate-limit window resets.
  - name: Token Reuse
    description: >-
      Cache OAuth2 JWT access tokens and reuse them until near-expiry (check the exp claim).
      Requesting a new token per API call can itself trigger rate limiting.
  - name: Response Caching
    description: >-
      Cache API responses for frequently-requested area/time combinations to reduce PU
      consumption and request counts.
maintainers:
  - FN: Kin Lane
    email: kin@apievangelist.com