Coresignal · Rate Limits

Coresignal Rate Limits

Coresignal enforces one rate limit per API key, set by the subscribed plan and expressed in requests per second. The same ceiling covers search, collect, bulk collect, search preview, webhook requests and the Agentic Search /fast endpoint; there is no separate budget per verb or per dataset. The Agentic Search /reasoning endpoint is the single exception, limited to 10 requests per hour. Credit allocation is a separate, orthogonal control - a monthly quota that governs how many records you may pull, not how fast you may ask.

Coresignal Rate Limits is the machine-readable rate-limit profile for Coresignal on the APIs.io network, conforming to the API Commons Rate Limits specification.

It captures 9 rate-limit definitions, measuring requests_per_second and requests_per_hour.

The profile also includes 4 backoff/retry policies defined and response codes documented for throttled and quotaExceeded.

Tagged areas include Agentic Search, B2B Data, Companies, Data as a Service, and Employees.

9 Limits Throttle: 429 Quota: 402
Agentic SearchB2B DataCompaniesData as a ServiceEmployeesJob PostingsRate LimitingQuotasThrottling

Limits

Data APIs - Free trial api-key
requests_per_second · second
5
Data APIs - Mini api-key
requests_per_second · second
5
Data APIs - Starter api-key
requests_per_second · second
5
Data APIs - Pro api-key
requests_per_second · second
10
Data APIs - Growth api-key
requests_per_second · second
20
Data APIs - Premium api-key
requests_per_second · second
50
Data APIs - Scale api-key
requests_per_second · second
100
Data APIs - Elite api-key
requests_per_second · second
100
Published as "100+ req/s"; the ceiling above 100 is not stated and is presumably negotiated.
Agentic Search API - /v2/agentic_search/reasoning endpoint
requests_per_hour · hour
10
Endpoint-specific and plan-independent. The /v2/agentic_search/fast endpoint is governed by the plan's per-second limit instead.

Policies

Per-API-key scoping
Limits are enforced per API key. Multiple keys can be generated per account, so issue separate keys per environment or downstream tenant for isolation - that is the only isolation mechanism, since keys carry no scopes.
Backoff strategy
On 429, back off and retry. No Retry-After is sent, so choose your own schedule; the window is per-second, so a short pause typically clears it.
Credits versus throttling
429 means too fast; 402 means out of credits. They are different controls and a 402 will not clear by waiting - it clears on renewal or top-up.
Search is free, so rehearse with it
/search/es_dsl and /search/filter deduct no credits. Use search to size a result set from x-total-results before committing to a collect that spends the budget.

Sources