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: Leonardo.AI
providerId: leonardo-ai
created: '2026-05-25'
modified: '2026-05-25'
reconciled: true
tags:
  - AI
  - Rate Limiting
  - Concurrency
  - Queue
description: Reconciled concurrency, queue, and rate-limit behaviour for the Leonardo.AI Production API.
  Leonardo enforces per-API-key concurrency limits (parallel in-flight generations) on top of a queueing system
  rather than a strict RPM/TPM token bucket. Excess requests are queued, not 429'd, up to the queue depth.
sources:
  - https://docs.leonardo.ai/docs/concurrency-rate-limits-and-queue
  - https://docs.leonardo.ai/docs/api-faq
  - https://docs.leonardo.ai/docs/api-error-messages
responseCodes:
  unauthorized: 401
  forbidden: 403
  notFound: 404
  validationError: 400
  rateLimited: 429
  serverError: 500
algorithm: per-key-concurrency-with-queue
notes:
  - Each Production API key has a concurrency limit — the maximum number of in-flight generations Leonardo will
    process in parallel for that key. Additional requests are queued, not rejected, up to the queue depth.
  - Concurrency and queue depth scale with account standing and historical usage. Contact Leonardo support to
    request an increased limit for production workloads.
  - Synchronous endpoints (Pricing Calculator, /me, list endpoints, prompt utilities) are subject to a
    separate, higher RPM-style ceiling and may return 429 if the caller floods them.
  - Webhook callbacks are the preferred completion-notification path; the FAQ explicitly recommends webhooks
    over polling GET /generations/{id}.
  - Up to 10 Production API keys may be issued per account.
limits:
  - tier: Default Production API
    model: All generation endpoints
    concurrency: variable
    queue: yes
    rpm: not-published
    notes: Concurrency and queue depth are account-scoped, not publicly documented. Use the in-app API
      Access dashboard and contact support for production limit increases.
  - tier: Synchronous utility endpoints
    model: "/pricing-calculator, /me, /platformModels, /prompt/*"
    concurrency: higher
    queue: no
    rpm: not-published
    notes: Designed for higher request rates than the generation endpoints, but still subject to a fair-use
      ceiling.
keyManagement:
  maxKeysPerAccount: 10
  rotation: Production API keys can be rotated from the in-app API Access dashboard. The legacy User API key
    is fully deprecated; all integrations must use a Production API key.