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: Semrush
providerId: semrush
created: '2026-05-04'
generated: '2026-08-13'
method: searched
source: https://developer.semrush.com/api/v4/introduction/api-usage-restrictions/
docs: https://developer.semrush.com/api/v4/introduction/api-usage-restrictions/
modified: '2026-08-13'
tags:
  - Data
  - Search Engines
  - Rate Limiting
  - Quotas
  - Throttling
description: >-
  Published rate limits for the Semrush API surface, read from the provider's API usage
  restrictions page. Replaces the 2026-05-04 bulk-sweep scaffold, whose header names and
  tier numbers were invented defaults. The real posture: a flat 10 requests per second and
  10 concurrent requests per account across every Semrush API, per-method exceptions on the
  Listing Management API, and — critically — no rate-limit response headers of any kind.

headers:
  limit: null
  remaining: null
  reset: null
  retryAfter: null
  policy: null
headers_note: >-
  Semrush documents no X-RateLimit-*, no RateLimit-* (draft IETF) and no Retry-After header
  on any API. The 2026-05-04 scaffold asserted all five; none of them appear in the provider's
  documentation. A client has no runtime budget signal and learns it is throttled only by
  receiving a 429. The one adjacent runtime signal is error.retryable in the error body.

responseCodes:
  throttled: 429
  quotaExceeded: 429
  serviceUnavailable: 503
  serviceUnavailableLocal: 553
responseCodes_note: >-
  The Local API remaps server-side failures into the 55x range (550/551/553/554), so backoff
  logic written against standard 5xx codes will not fire there. See errors/semrush-problem-types.yml.

limits:
  - name: General request rate
    scope: account
    metric: requests_per_second
    limit: 10
    timeFrame: second
    applies:
      - All Semrush API methods
    source: https://developer.semrush.com/api/v4/introduction/api-usage-restrictions/
  - name: General concurrency
    scope: account
    metric: concurrent_requests
    limit: 10
    timeFrame: instant
    applies:
      - All Semrush API methods
    source: https://developer.semrush.com/api/v4/introduction/api-usage-restrictions/
  - name: Trends API request rate
    scope: account
    metric: requests_per_second
    limit: 10
    timeFrame: second
    applies:
      - Trends API (v3)
    note: >-
      Daily and hourly limits were removed. Those fields remain in the API response for
      backward compatibility but their values are no longer used or decremented — a stale
      field that will mislead any client still reading it.
  - name: Trends API monthly quota
    scope: account
    metric: requests_per_month
    limit: null
    timeFrame: month
    applies:
      - Trends API (v3)
    note: >-
      Subscription-dependent. The default monthly limit included with Trends API access is
      10,000 requests; the whole monthly quota may be spent at any pace within the 10 RPS cap.
  - name: Create Location
    scope: account
    metric: requests_per_second
    limit: 10
    timeFrame: second
    applies:
      - Listing Management API v4
  - name: Update Location
    scope: account
    metric: requests_per_second
    limit: 10
    timeFrame: second
    applies:
      - Listing Management API v4
  - name: Get Location
    scope: account
    metric: requests_per_second
    limit: 10
    timeFrame: second
    applies:
      - Listing Management API v4
  - name: Get Locations
    scope: account
    metric: requests_per_second
    limit: 10
    timeFrame: second
    applies:
      - Listing Management API v4
  - name: Get Categories
    scope: account
    metric: requests_per_minute
    limit: 10
    timeFrame: minute
    applies:
      - Listing Management API v4
    note: The one endpoint limited per minute rather than per second — 600x tighter than its neighbours.
  - name: UpdateLocation (deprecated)
    scope: account
    metric: requests_per_second
    limit: 5
    timeFrame: second
    applies:
      - Listing Management API (OAuth 2.0, deprecated)
  - name: UpdateLocations (deprecated)
    scope: account
    metric: requests_per_minute
    limit: 5
    timeFrame: minute
    applies:
      - Listing Management API (OAuth 2.0, deprecated)
    note: Up to 50 locations per request.
  - name: GetLocation (deprecated)
    scope: account
    metric: requests_per_second
    limit: 10
    timeFrame: second
    applies:
      - Listing Management API (OAuth 2.0, deprecated)
  - name: GetLocations (deprecated)
    scope: account
    metric: requests_per_second
    limit: 10
    timeFrame: second
    applies:
      - Listing Management API (OAuth 2.0, deprecated)

limit_count: 13

consumption:
  model: api-units
  note: >-
    Beyond request rate, Standard API usage is metered in API units billed per returned line
    and weighted by report type — one line of Domain Organic Search Keywords costs 10 units
    live and 50 units historical. display_limit caps returned lines and is therefore the
    main cost control. Balance is readable free at
    http://www.semrush.com/users/countapiunits.html?key=<key>.
  mcp_note: The MCP server consumes the same API unit balance.

other_restrictions:
  - name: Caching
    detail: >-
      Information received via the API may not be cached for more than one month without
      Semrush's express written consent (Terms of Service section 3.3).

checked: '2026-08-13'