generated: '2026-10-07'
method: searched
source: https://docs.mrscraper.com/docs/api/overview
sources:
- https://docs.mrscraper.com/docs/api/authentication
- https://docs.mrscraper.com/docs/api/pagination
- https://docs.mrscraper.com/docs/api/response-headers
- https://docs.mrscraper.com/docs/api/error-handling
- https://docs.mrscraper.com/docs/api/v3/scraper/bulk/cancel-bulk-by-result-id
- openapi/mrscraper-scraper-service-gateway-openapi.yml
authentication:
  style: api-key header
  header: x-api-token
  alternatives: 'Authorization: Bearer <API token> on Marketplace endpoints; `token` query parameter on the Scraper API; OAuth 2.1 bearer on the hosted MCP endpoint'
  see: authentication/mrscraper-authentication.yml
idempotency:
  coverage: none
  header: null
  note: No Idempotency-Key header or replay protection is documented for any write (POST /api/v1/scrapers-ai, reruns, bulk reruns, Scraper API POST /). The Scraper API `async` flag returns a task ID immediately but does not de-duplicate repeated submissions.
reversibility:
  grade: documented
  write_surface:
  - operation: Bulk rerun (POST /api/v1/scrapers-ai-rerun/bulk, POST /api/v1/scrapers-manual-rerun/bulk)
    reversal: PATCH /api/v1/scrapers-bulks/result/{resultId}/cancel (Cancel One Bulk Result by Result ID) and PATCH /api/v1/scrapers-bulks/{scraperId}/cancel (Cancel All Bulk Results by Scraper ID)
    window: not stated - the docs describe cancelling queued/running bulk results but publish no time window
    docs: https://docs.mrscraper.com/docs/api/v3/scraper/bulk/cancel-bulk-by-scraper-id
  - operation: Single scraper runs (POST /api/v1/scrapers-ai, /scrapers-ai-rerun, /scrapers-manual-rerun, Scraper API POST /)
    reversal: none documented
    window: null
  dry_run_mode: none documented
  note: Token consumption from a completed run is not documented as refundable.
pagination:
  style: page-number
  params: [page, pageSize, sortField, sortOrder]
  optional_params: [search, dateRangeColumn, startAt, endAt]
  response_fields: [meta.page, meta.pageSize, meta.total, meta.totalPage]
  max_page_size: 100 (recommended maximum to avoid timeouts)
  note: The pagination page calls the scheme cursor-based, but every documented parameter is page/pageSize with a total count. The gateway /jobs and /tasks list endpoints take the same parameters.
field_expansion: none documented
metadata: none documented
request_tracing:
  response_headers: [x-result-id]
  usage_headers: [token_usage, bandwidth_usage, runtime, x-status-code]
  note: Usage headers are returned only by the Manual Scraper API and Web Renderer API; x-status-code is the TARGET site's HTTP status, x-result-id identifies the stored result.
versioning:
  scheme: path (/api/v1 on api.app.mrscraper.com); Scraper API gateway unversioned
  see: lifecycle/mrscraper-lifecycle.yml
error_envelope:
  shape: '{"message": string, "error": string}'
  content_type: application/json
  see: errors/mrscraper-problem-types.yml
rate_limit_signaling:
  status: 429
  headers: [Retry-After]
  note: Retry-After is described as "if present"; no X-RateLimit-* or RateLimit headers are documented.
  see: rate-limits/mrscraper-rate-limits.yml
content_type: application/json for every request with a payload and every response (binary screenshots return base64 inside JSON)