generated: '2026-08-13'
method: searched
source: https://developers.similarweb.com/docs/rate-limit
also_documented_at:
- https://docs.similarweb.com/api-v5/guides/error-handling-and-troubleshooting
limit_count: 3
name: SimilarWeb API Rate Limits
description: >-
  SimilarWeb enforces rate limits on the REST API to ensure platform stability
  and fair use across all customers. The primary limit is 10 requests per second
  per API key. The Batch API is available for high-volume use cases and processes
  asynchronously, bypassing per-second rate limits. Enterprise customers may
  negotiate higher rate limits with the SimilarWeb team.
specificationVersion: '0.1'
url: https://developers.similarweb.com/docs/rate-limit

limits:
  - name: Requests Per Second
    description: >-
      The REST API is capped at 10 requests per second per API key. Exceeding
      this limit returns HTTP 429 Too Many Requests. No data credits are consumed
      when a request is rate-limited.
    type: rate
    scope: api-key
    value: 10
    unit: requests
    period: second
    errorCode: 429
    errorMessage: Too many requests
    creditsConsumed: false

  - name: Monthly Data Credit Quota
    description: >-
      Each API plan includes a monthly allocation of data credits. Credits are
      consumed per API call based on the endpoint accessed and the volume of data
      returned. The quota resets monthly. Customers can use the Request Validate
      endpoint to estimate credit costs before running large queries.
    type: quota
    scope: account
    unit: data-credits
    period: monthly
    resetBehavior: monthly
    validateEndpoint: true
    notes: >-
      Credit costs vary by endpoint. High-value endpoints (e.g., keyword
      intelligence, app performance) typically consume more credits per call
      than basic traffic summary endpoints.

  - name: Batch API Job Size
    description: >-
      Individual Batch API jobs can include up to one million domains. Batch
      jobs are processed asynchronously and do not count against the per-second
      rate limit. Results are delivered to Amazon S3, Google Cloud Storage,
      or Snowflake.
    type: size
    scope: job
    value: 1000000
    unit: domains
    period: per-job
    delivery:
      - Amazon S3
      - Google Cloud Storage
      - Snowflake

headers:
  published: []
  retry_after: false
  observed_response_headers:
    - date
    - server
    - strict-transport-security
    - x-cache
    - via
    - x-amz-cf-id
    - x-amz-cf-pop
  note: >-
    Similarweb emits NO rate-limit response headers. There is no X-RateLimit-Limit,
    X-RateLimit-Remaining, X-RateLimit-Reset, no RFC 9239-style RateLimit-* header, and no
    Retry-After on a 429. Verified against the docs and against a live unauthenticated
    request to api.similarweb.com on 2026-08-13, whose response carried only the headers
    listed above (CloudFront edge headers plus HSTS). This is the practical gap for an
    agent: the limit is knowable only from documentation, never from the runtime response,
    so a client must count requests itself and cannot discover its remaining budget.
  statuses:
    - name: HTTP 429
      description: Returned when the 10 requests-per-second limit is exceeded (REST).
      action: Implement client-side throttling; retry after a brief back-off period.
      credits_consumed: false
    - name: HTTP 429 (Batch)
      description: >-
        On the Batch API, 429 means the account exceeded its limit of allowed PENDING
        requests — a concurrency limit on in-flight jobs, not a per-second limit.
      action: Let queued jobs complete before submitting more; do not simply slow the request rate.

x-evidence:
  fetched: '2026-08-13'
  probes:
    - url: https://developers.similarweb.com/docs/rate-limit.md
      http_status: 200
    - url: 'https://api.similarweb.com/v1/website/cnn.com/total-traffic-and-engagement/visits (invalid key)'
      http_status: 401
      rate_limit_headers_present: false

bestPractices:
  - Implement client-side throttling to stay at or below 10 requests per second.
  - Use the Request Validate endpoint before running large or expensive queries.
  - Use the Batch API for bulk domain analysis rather than looping REST calls.
  - Monitor monthly credit consumption to avoid exhausting quotas mid-month.
  - Contact SimilarWeb support to discuss higher rate limits for enterprise workloads.