generated: '2026-08-13'
method: searched
source: https://www.demandsphere.com/platform/apis/rest-apis/
http_status: 200
checked: '2026-08-13'
summary: >-
  DemandSphere publishes no numeric rate limit for its REST API. The developer
  page states, verbatim, that the REST APIs come with "no artificial rate
  limits", and neither the page nor the OpenAPI declares a 429 response, a
  RateLimit-* / X-RateLimit-* header, or a Retry-After contract. The API does
  return 429 in practice: DemandSphere's own first-party MCP server treats 429
  as a retryable status and classifies it as `rate_limited`, which is the only
  provider-published evidence of an enforced ceiling.
limit_count: 0
published_limits: []
provider_statement:
  quote: >-
    REST APIs for SERP rankings, LLM visibility, citations, log analytics, and
    more. OAuth 2.0 authentication, comprehensive documentation, and no
    artificial rate limits.
  url: https://www.demandsphere.com/platform/apis/rest-apis/
response_headers:
  documented: false
  observed: []
  note: >-
    No RateLimit-*, X-RateLimit-* or Retry-After header is documented. The
    published OpenAPI declares only 200 and 400 on every operation, so an agent
    has no declared runtime signal to read.
exhaustion:
  status_code: 429
  documented_in_openapi: false
  evidence:
    source: https://github.com/DemandSphereDev/demandsphere-mcp
    file: src/demandsphere_mcp/client.py
    detail: >-
      `_RETRYABLE_STATUS = {429, 500, 502, 503, 504}` — the first-party client
      retries 429 with backoff; tools/utils.py maps status 429 to error_type
      `rate_limited` with the recovery hint "Rate limit reached. Wait a moment
      and retry".
client_side_limits:
- scope: per MCP server process
  limit: 60
  window: minute
  mechanism: token bucket
  configurable: DEMANDSPHERE_MAX_REQUESTS_PER_MINUTE
  source: https://github.com/DemandSphereDev/demandsphere-mcp
  note: >-
    A self-imposed client default in DemandSphere's own MCP server, not a
    server-enforced quota. Recorded because it is the only number the provider
    publishes anywhere near a rate limit.
- scope: per MCP tool call
  limit: 100
  unit: rows returned
  configurable: DEMANDSPHERE_MAX_RESULTS_PER_TOOL_CALL
  source: https://github.com/DemandSphereDev/demandsphere-mcp
  note: Response shaping cap, not a request-rate limit.
quota:
  model: subscription volume, not request rate
  note: >-
    Commercial ceilings are expressed as tracked keyword/prompt volume per plan
    (Starter 1,000 keywords / 100 prompts; Pro 2,500 / 250; Enterprise custom),
    not as API calls. See plans/demandsphere-plans-pricing.yml.
cross_links:
  plans: plans/demandsphere-plans-pricing.yml
  conventions: conventions/demandsphere-conventions.yml
  errors: errors/demandsphere-problem-types.yml