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: Jina AI
providerId: jina-ai
created: '2026-05-04'
# Provenance stamped 2026-08-11: this artifact was written by the API Evangelist
# bulk sweep dated 2026-05-04, not harvested from the provider. See roadmap#35.
method: generated
modified: '2026-05-05'
reconciled: true
tags:
  - Rate Limiting
  - AI
  - Embeddings
  - LLM
description: Jina AI applies per-API-key rate limits across three tiers (Free, Paid, Premium). Limits
  are enforced as RPM (requests per minute), TPM (tokens per minute), and concurrent in-flight requests
  and apply uniformly across all Search Foundation services (Embeddings, Reranker, Reader, Classifier,
  Segmenter, DeepSearch). Tier is determined by token balance / billing status on the key.
sources:
  - https://jina.ai/embeddings/
  - https://jina.ai/api-dashboard/
  - https://api.jina.ai/redoc
headers:
  retryAfter: Retry-After
responseCodes:
  throttled: 429
  quotaExceeded: 402
limits:
  - name: Free API Key — RPM
    scope: api-key
    metric: requests_per_minute
    limit: 100
    timeFrame: minute
  - name: Free API Key — TPM
    scope: api-key
    metric: tokens_per_minute
    limit: 100000
    timeFrame: minute
  - name: Free API Key — concurrency
    scope: api-key
    metric: concurrent_requests
    limit: 2
  - name: Paid API Key — RPM
    scope: api-key
    metric: requests_per_minute
    limit: 500
    timeFrame: minute
  - name: Paid API Key — TPM
    scope: api-key
    metric: tokens_per_minute
    limit: 2000000
    timeFrame: minute
  - name: Paid API Key — concurrency
    scope: api-key
    metric: concurrent_requests
    limit: 50
  - name: Premium API Key — RPM
    scope: api-key
    metric: requests_per_minute
    limit: 5000
    timeFrame: minute
  - name: Premium API Key — TPM
    scope: api-key
    metric: tokens_per_minute
    limit: 50000000
    timeFrame: minute
  - name: Premium API Key — concurrency
    scope: api-key
    metric: concurrent_requests
    limit: 500
policies:
  - name: Tier upgrade
    description: Tier (Free / Paid / Premium) is determined by token balance / billing status on the
      API key; topping up tokens promotes the key to a higher tier with higher RPM, TPM, and concurrency.
  - name: Backoff
    description: Honor Retry-After on 429 and back off exponentially with jitter.
  - name: Token exhaustion
    description: When the token balance reaches zero the API returns 402 Payment Required; top up
      via Stripe in the API Dashboard.
  - name: Shared across services
    description: Rate-limit and concurrency caps apply across all Jina services on the same key, not
      per-endpoint.
maintainers:
  - FN: API Evangelist
    email: info@apievangelist.com