Google Gemini API 503 UNAVAILABLE: what it means and how to fix it

503 UNAVAILABLE: The model is overloaded. Please try again later.

How to fix it

1. Retry with exponential backoff and jitter.
2. Fall back to another Gemini model (a Flash model) or another provider for user-facing requests.

Is it down right now?

Check our live status page or the official one.

Retry it properly

import random, time
from google import genai
from google.genai import errors

client = genai.Client()

def generate(prompt, tries=6):
    for i in range(tries):
        try:
            return client.models.generate_content(model="gemini-flash-latest", contents=prompt)
        except errors.APIError as e:
            # 429 RESOURCE_EXHAUSTED, 503 UNAVAILABLE, 500, 504: transient. 400/402/403: fix, do not retry.
            if e.code not in (429, 500, 503, 504) or i == tries - 1:
                raise
            time.sleep(min(60, 2 ** i) + random.random())  # exponential backoff + jitter

Google lists 429 RESOURCE_EXHAUSTED, 503 UNAVAILABLE, 408 and 5xx as transient: retry with exponential backoff and jitter. 400, 402 and 403 are not.

Same kind of error at other providers

• Anthropic 529 overloaded_error
• Anthropic 500 api_error
• Anthropic 504 timeout_error
• OpenAI 500 server_error

Other Google Gemini API errors

• 429RESOURCE_EXHAUSTEDretry
• 400INVALID_ARGUMENTfix
• 400FAILED_PRECONDITIONfix
• 402payment_requiredfix
• 403PERMISSION_DENIED: Your project has been denied accessfix
• 403PERMISSION_DENIED: Your API key was reported as leaked. Please use another API key.fix
• 404NOT_FOUNDfix
• 500INTERNALretry
• 504DEADLINE_EXCEEDEDretry
• -RECITATIONfix
• -SAFETYfix

Source

• Google Gemini API error documentation: ai.google.dev/gemini-api/docs/api-errors (checked Oct 24, 2026)

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