Deepchecks · Rate Limits

Deepchecks Rate Limits

The Deepchecks LLM Evaluation cloud API does not publish explicit numeric request-per-second/minute rate limits. Practical capacity is governed by plan entitlements - number of AI applications, seats, and DPUs (Data Processing Units) per month - rather than by published HTTP throttling thresholds. Interaction logging is designed for batched and streamed ingestion. The open-source packages run locally and impose no API rate limits. Specific per-endpoint limits are not reconciled in this artifact.

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

It captures 4 rate-limit definitions, measuring dpus, applications, seats, and requests.

The profile also includes 2 backoff/retry policies defined and response codes documented for throttled.

Tagged areas include AI, LLM, Evaluation, Testing, and Monitoring.

4 Limits Throttle: 429
AILLMEvaluationTestingMonitoringRate LimitingQuotasThrottling

Limits

DPUs Per Month account
dpus
see plan (5K Basic / 20K Scale / custom Enterprise)
Data Processing Units consumed by evaluation; primary capacity governor per plan.
AI Applications account
applications
see plan (1 Basic / 3 Scale / custom Enterprise)
Number of concurrently configured evaluation applications.
Seats account
seats
see plan (3 Basic / 5 Scale / custom Enterprise)
Number of platform user seats.
Interaction Ingestion account
requests
see provider documentation
No published RPM/RPS; batch and streaming ingestion recommended for high volume.

Policies

Plan-Governed Capacity
Throughput is bounded by DPU, application, and seat entitlements rather than published HTTP rate limits.
Backoff Strategy
Clients should implement exponential backoff with jitter and honor Retry-After on 429 responses.

Sources