Spark Platform Rate Limits
Spark Platform publishes no numeric request-rate limit. What it publishes is the throttle signal (HTTP 429 with Spark error code 1550 "Over rate limit") and a set of hard per-request ceilings on the pagination parameters, which are the only numbers FBS states in the open. The request-size ceilings differ by key role: a standard key may ask for at most 25 records per request, a replication-role key up to 1000. Long-running requests are cancelled server-side rather than queued (codes 1515 and 1516). No RateLimit-* or X-RateLimit-* response headers are documented, so a client cannot see its remaining budget — it learns by being refused.
Spark Platform Rate Limits is the machine-readable rate-limit profile for Spark Platform on the APIs.io network, in the API Commons Rate Limits shape.
It captures 4 rate-limit definitions, measuring records_per_request, page_number, and record_offset.
Limit state is signalled in the rateLimitRemaining, retryAfter, and note response headers, and an exhausted limit returns HTTP 429.
The profile also includes response codes documented for throttled, throttledProviderCode, timeoutCancelled, timeoutProviderCodes, temporarilyUnavailable, and temporarilyUnavailableProviderCode.
Tagged areas include Rate Limiting, MLS, and Real Estate.
rateLimitRemainingretryAfternote
Limits
Sources
- https://sparkplatform.com/docs/supporting_documentation/error_codes
- https://sparkplatform.com/docs/supporting_documentation/search_and_paging_syntax
- https://sparkplatform.com/docs/supporting_documentation/spark_replication
Work with this as data
Every rate limit here is available over the APIs.io API and to AI agents over MCP.