Hosted LLM providers like OpenAI and Anthropic have rate limiting capabilities centered around requests per minute (RPM) and tokens per minute (TPM). These are specified at the “organization” level and “project” level. API keys are associated with an organization or pro