What happens when a rate limit is exceeded?

The API returns an HTTP 429 (Too Many Requests) status code with a JSON error body describing which limit was hit. Most official SDKs (OpenAI Python, Anthropic Python) include built-in retry logic that handles 429 errors automatically with exponential backoff. For custom implementations, inspect the

Do rate limits apply per API key or per organization?

This varies by provider. OpenAI and Anthropic enforce rate limits at the organization level, meaning all API keys within the same organization share the same rate limit budget. Creating additional API keys within the same org does not increase total capacity. Google Gemini enforces limits per projec

Can rate limits be increased without upgrading tiers?

Yes, in some cases. OpenAI and Anthropic both offer the ability to request rate limit increases for specific use cases by contacting their sales or support teams. Enterprise plans from all major providers include custom rate limits negotiated based on expected usage. For Google, Tier 3 enterprise li

How do cached tokens affect rate limits?

Cached tokens provide a significant advantage for rate limits. With Anthropic, only uncached input tokens count toward ITPM rate limits, effectively increasing throughput when prompt caching is active. OpenAI's cached input tokens are processed faster and do not count against standard TPM limits in

AI Tools Kit

API & Pricing

AI API Rate Limits Explained: OpenAI, Anthropic, Google & More (2026)

Complete guide to AI API rate limits across all major providers. How rate limits work, current limits by tier, and strategies to avoid hitting them.

Published March 7, 2026

12 min read

AI Tools Kit

AI Tools Kit provides free developer tools for working with AI language models. Built by developers, for developers.

Learn more about us →