OpenAI API error 429 'insufficient_quota': 'You exceeded your current quota, please check your plan and billing details' even on my first request. How do I fix it?

Cause: this 429 is a billing error, not a rate limit. The API is prepaid: the organization or project the key belongs to has no remaining credit, or it reached the monthly budget / usage limit set for it. A ChatGPT Plus or Pro subscription does not include API credit, which is why brand-new keys often fail on the first call. It is different from 429 'rate_limit_exceeded' (too many requests or tokens per minute), which goes away by waiting and retrying. How to check: - Read the error body: code 'insufficient_quota' means billing; code 'rate_limit_exceeded' means rate limits. - In the OpenAI platform, open Settings, Billing for the organization the key belongs to and check the credit balance, and check Limits (organization and project budgets). - Make sure the key belongs to the organization and project you funded. Keys from another org or a project with a budget of zero fail even when your main account has credit. Fix: 1. Add prepaid credit to the organization (or enable auto-recharge), and raise the organization or project budget if it was reached. 2. If you just added credit, wait a few minutes; if it still fails, create a new API key in the funded project and use that. 3. Don't retry insufficient_quota in a loop: it will keep failing until billing changes. Have your code stop and report it, and retry only rate_limit_exceeded with backoff. After: set a usage alert and a sensible auto-recharge or budget so long-running agents don't stop silently when credit runs out. Sources: https://platform.openai.com/docs/guides/error-codes https://help.openai.com/en/articles/6891831-error-code-429-you-exceeded-your-current-quota-please-check-your-plan-and-billing-details https://community.openai.com/t/429-error-insufficient-quota/492350