OpenAI 429 insufficient_quota — It's Billing, Not a Rate Limit

Before you change anything, read the error.code in the 429 response body. In the official OpenAI SDKs, both insufficient_quota and rate_limit_exceeded raise RateLimitError — so you must branch on the code, not the exception type. Stop on billing; back off on rate limits:

# Python — branch on error.code, not just the exception type
from openai import OpenAI, RateLimitError

client = OpenAI()  # reads OPENAI_API_KEY from the environment

try:
    resp = client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": "ping"}],
    )
except RateLimitError as e:
    # IMPORTANT: insufficient_quota AND rate_limit_exceeded both raise
    # RateLimitError with HTTP 429. Only error.code tells them apart.
    code = getattr(e, "code", None)
    if code == "insufficient_quota":
        # BILLING problem — retrying / backoff will NEVER fix this.
        raise SystemExit("No API credit. Add a payment method + buy credits.")
    elif code == "rate_limit_exceeded":
        # Real throughput limit — THIS one you back off and retry.
        print("Rate limited — sleep, then retry with exponential backoff.")
    else:
        raise

// TypeScript — same branch on e.code
import OpenAI from 'openai';

const client = new OpenAI(); // reads process.env.OPENAI_API_KEY

try {
  await client.chat.completions.create({
    model: 'gpt-4o-mini',
    messages: [{ role: 'user', content: 'ping' }],
  });
} catch (e) {
  if (e instanceof OpenAI.RateLimitError) {
    // 429 covers BOTH billing and true rate limits — check the code:
    if (e.code === 'insufficient_quota') {
      // BILLING problem — backoff will NOT help. Stop and add credits.
      throw new Error('No API credit. Add a payment method + buy credits.');
    } else if (e.code === 'rate_limit_exceeded') {
      console.warn('Rate limited — back off and retry.');
    }
  }
}

This fixes the vast majority of cases. Open the billing overview and look at your available balance:

If your account only ever had free-trial credits, those expire after roughly 3 months and cannot be reused, extended, or reactivated. Once expired, the balance shows $0 and every request returns insufficient_quota.

There is no workaround here — you must add a paid payment method and purchase credits. Trial credit is a one-time grant, not a recurring allowance.

Even with credit available, a hard monthly cap you (or an org admin) set earlier will return insufficient_quota once you cross it. This is easy to miss on organizations with multiple projects.

After adding credit, allow a few minutes for billing to propagate — the change is not always instant. Then send a tiny test request. A successful response confirms your quota is live:

# Python — minimal test call after topping up credit
from openai import OpenAI

client = OpenAI()
r = client.chat.completions.create(
    model="gpt-4o-mini",
    messages=[{"role": "user", "content": "say ok"}],
    max_tokens=5,
)
print(r.choices[0].message.content)  # prints "ok" once credit is live

// TypeScript — minimal test call
import OpenAI from 'openai';

const client = new OpenAI();
const r = await client.chat.completions.create({
  model: 'gpt-4o-mini',
  messages: [{ role: 'user', content: 'say ok' }],
  max_tokens: 5,
});
console.log(r.choices[0].message.content); // "ok" => quota is active

Still getting insufficient_quota after 10–15 minutes? Re-confirm you actually purchased credits (not just added a card), and that the key belongs to the project you funded.