To maintain consistent performance and availability, the APIs apply global, API-specific, and endpoint-specific rate limits.
All limits are measured per Company. Requests made with different API keys belonging to the same Company share the Company’s limits. The most restrictive applicable limit takes precedence.
How rate limiting works
Snappy uses a token bucket model. Each bucket has two parameters:
- Sustained rate - the rate at which tokens are added to the bucket, measured in requests per second (RPS). Each request, read or write, consumes one token.
- Burst capacity - the maximum number of tokens the bucket can hold, determining how many requests can be handled in a short burst.
When traffic stays below the sustained rate, unused tokens accumulate up to the burst capacity. These saved tokens provide temporary headroom for traffic spikes. After they are consumed, requests must follow the sustained rate until the bucket refills.
API rate limits
Endpoint-specific limits
Endpoint-specific export limits
Rate-limit responses
When a rate limit is exceeded, the API returns a 429 Too Many Requests response:
Clients should throttle requests according to the documented rate and burst limits. After receiving a 429 Too Many Requests response, clients should reduce their request rate before retrying. Requests that continue to exceed the applicable limit may receive additional 429 responses.
Fair-use protection
Additional temporary safeguards may be applied when traffic patterns threaten platform stability, including unusually large bursts, excessive polling, repeated failures, or automated abuse. Clients affected by these safeguards receive a 429 Too Many Requests response.
Higher limits may be approved for verified integrations with demonstrated business requirements. Last modified on July 27, 2026
