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: Klu
providerId: klu-ai
created: '2026-06-20'
modified: '2026-06-20'
reconciled: false
tags:
- AI
- LLM
- LLM App Platform
- Rate Limiting
- Quotas
- Throttling
description: >-
  Klu does not publish explicit numeric rate limits for the Klu Engine API in its
  public documentation. Generation throughput is effectively bounded by the
  upstream LLM provider limits (OpenAI, Anthropic, and others) tied to the keys
  connected to a workspace, and by any account-level limits Klu applies to Action
  runs. Specific per-account or per-endpoint values are not reconciled in this
  artifact.
notes: >-
  reconciled false - no documented RPM/TPM figures were found on docs.klu.ai.
  Clients should expect provider-side throttling on upstream model calls and Klu
  account limits on Action runs. Verify against the workspace settings and current
  documentation on reconciliation; note the product appears dormant (see review.yml).
sources:
- https://docs.klu.ai/resources/api-basics
- https://docs.klu.ai/resources/api-actions
- https://klu.ai
responseCodes:
  throttled: 429
limits:
- name: Action Runs
  scope: account
  metric: requests
  limit: see provider documentation
  notes: No published numeric limit on POST /v1/actions; bounded by account and upstream provider limits.
- name: Upstream Model Tokens
  scope: account
  metric: tokens
  limit: see upstream provider documentation
  notes: Token throughput is governed by the connected LLM provider's rate limits for the workspace key.
- name: Context Ingestion
  scope: account
  metric: requests
  limit: see provider documentation
  notes: Document and file ingestion into Context libraries; no published numeric limit.
policies:
- name: Upstream Pass-Through
  description: Generation limits largely follow the rate limits of the upstream LLM provider connected to the workspace.
- name: Backoff Strategy
  description: Clients should implement exponential backoff with jitter and honor 429 / Retry-After responses.
maintainers:
- FN: Kin Lane
  email: kin@apievangelist.com