OpenAI 503 server_is_overloaded: what it means and how to fix it

503 server_is_overloaded

How to fix it

1. Honor Retry-After if present, then retry with backoff.
2. Fall back to another model for user-facing requests.

Is it down right now?

Check our live status page or the official one.

Retry it properly

from openai import OpenAI, RateLimitError, InternalServerError

# The SDK retries some errors with backoff; quota and billing errors never clear by retrying.
client = OpenAI(max_retries=6)

try:
    r = client.responses.create(model="gpt-5", input="Hello")
except RateLimitError as e:          # 429: check e.code for quota vs rate limit
    if getattr(e, "code", None) in ("insufficient_quota", "credit_balance_exhausted"):
        raise  # fix billing, do not retry
except InternalServerError:          # 500 / 503 server_is_overloaded
    pass  # retry later or fall back to another model

The official SDKs retry some errors automatically (max_retries); billing and quota errors never clear by retrying.

Same kind of error at other providers

• Anthropic 529 overloaded_error
• Anthropic 500 api_error
• Anthropic 504 timeout_error
• Google Gemini API 503 UNAVAILABLE

Other OpenAI errors

• 429rate_limit_error: Rate limit reached for requestsretry
• 429slow_down: Slow downretry
• 429insufficient_quota: You exceeded your current quota, please check your plan and billing defix
• 401invalid_api_key: Incorrect API key providedfix
• 401invalid_authentication: Invalid Authenticationfix
• 401ip_not_authorized: IP not authorizedfix
• 401You must be a member of an organization to use the APIfix
• 403unsupported_country_region_territory: Country, region, or territory not supportedfix
• 500server_error: The server had an error while processing your requestretry
• 400invalid_request_errorfix

Source

• OpenAI error documentation: developers.openai.com/api/docs/guides/error-codes (checked Oct 24, 2026)

Paying too much per call while you are here? Compare AI model prices or estimate your bill.