specification: API Commons Rate Limits
specificationVersion: '0.1'
schema: https://raw.githubusercontent.com/api-evangelist/interface-research/main/schema/api-commons.yml#/$defs/RateLimits
provider: Deepchecks
providerId: deepchecks
created: '2026-06-20'
modified: '2026-06-20'
reconciled: false
tags:
- AI
- LLM
- Evaluation
- Testing
- Monitoring
- Rate Limiting
- Quotas
- Throttling
description: >-
  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.
notes: >-
  Verify any per-account or per-endpoint throttling and DPU consumption rules in
  the Deepchecks LLM Evaluation documentation and account console on reconciliation.
  DPU monthly allotments effectively cap evaluation throughput per plan.
sources:
- https://llmdocs.deepchecks.com/reference
- https://www.deepchecks.com/pricing/
- https://llmdocs.deepchecks.com/docs/deepchecks-sdk
responseCodes:
  throttled: 429
limits:
- name: DPUs Per Month
  scope: account
  metric: dpus
  limit: see plan (5K Basic / 20K Scale / custom Enterprise)
  notes: Data Processing Units consumed by evaluation; primary capacity governor per plan.
- name: AI Applications
  scope: account
  metric: applications
  limit: see plan (1 Basic / 3 Scale / custom Enterprise)
  notes: Number of concurrently configured evaluation applications.
- name: Seats
  scope: account
  metric: seats
  limit: see plan (3 Basic / 5 Scale / custom Enterprise)
  notes: Number of platform user seats.
- name: Interaction Ingestion
  scope: account
  metric: requests
  limit: see provider documentation
  notes: No published RPM/RPS; batch and streaming ingestion recommended for high volume.
policies:
- name: Plan-Governed Capacity
  description: Throughput is bounded by DPU, application, and seat entitlements rather than published HTTP rate limits.
- name: Backoff Strategy
  description: Clients should implement exponential backoff with jitter and honor Retry-After on 429 responses.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com