# Rate limits transcribed from the provider's own documentation — see provenance. Quoted, not derived.
generated: '2026-09-20'
method: searched
source: https://portkey.ai/docs/guides/getting-started/tackling-rate-limiting
sources:
- https://portkey.ai/docs/guides/getting-started/tackling-rate-limiting
provenance:
  tool: planning/api-economics/harvest_ratelimits.py
  run: 20260920T133210
  grounding: every limit_text was found verbatim in the fetched page text
published: true
note: Cookbook page about avoiding rate limits with Portkey fallbacks/load balancing; the numeric limits
  listed are third-party LLM provider limits (OpenAI, Anthropic, Cohere, Anyscale, Perplexity, Together
  AI), not Portkey's own. Page also mentions tts-1-hd 'you can not send more than 7 requests in minute'
  and that 429 is generated for rate limit errors.
limit_count: 11
limits:
- name: OpenAI gpt-5 requests per minute (Tier 1)
  limit_text: 500 Requests per Minute
  limit: 500
  window: minute
  scope: account
  plan: Tier 1
  endpoint: gpt-5
- name: OpenAI gpt-5 tokens per minute (Tier 1)
  limit_text: 10,000 Tokens per Minute
  limit: 10000
  window: minute
  scope: account
  plan: Tier 1
  endpoint: gpt-5
- name: OpenAI gpt-5 requests per day (Tier 1)
  limit_text: 10,000 Requests per Day
  limit: 10000
  window: day
  scope: account
  plan: Tier 1
  endpoint: gpt-5
- name: Anthropic RPM (Tier 1)
  limit_text: 50 RPM
  limit: 50
  window: minute
  scope: account
  plan: Tier 1
  endpoint: All models
- name: Anthropic TPM (Tier 1)
  limit_text: 50,000 TPM
  limit: 50000
  window: minute
  scope: account
  plan: Tier 1
  endpoint: All models
- name: Anthropic tokens per day (Tier 1)
  limit_text: 1 Million Tokens per Day
  window: day
  scope: account
  plan: Tier 1
  endpoint: All models
- name: Cohere production key RPM
  limit_text: 10,000 RPM
  limit: 10000
  window: minute
  scope: api-key
  plan: Production Key
  endpoint: Co.Generate models
- name: Anyscale concurrency
  limit_text: 30 concurrent requests
  limit: 30
  scope: endpoint
  endpoint: All models
- name: Perplexity RPM
  limit_text: 24 RPM
  limit: 24
  window: minute
  scope: account
  endpoint: mixtral-8x7b-instruct
- name: Perplexity TPM
  limit_text: 16,000 TPM
  limit: 16000
  window: minute
  scope: account
  endpoint: mixtral-8x7b-instruct
- name: Together AI paid RPM
  limit_text: 100 RPM
  limit: 100
  window: minute
  scope: account
  plan: Paid
  endpoint: All models
exceeded_status: 429