What this OpenAI API error means, whether it is on OpenAI’s side or yours, and what to do now. All times are UTC.
Server-side: OpenAI’s capacity
This error comes from OpenAI’s servers, not from your code. If the status below shows a problem, OpenAI has confirmed it. If it shows Operational, the problem may be short-lived, or the status page may not show it yet; our own check is a second opinion.
OpenAI’s docs describe a 503 with type service_unavailable_error and code server_is_overloaded as the requested model not having enough capacity to process your request at the moment.
It is about the model’s capacity, not your account. OpenAI’s Python library raises it as InternalServerError, not RateLimitError, so handle both.
Retry-After header is present, wait at least that long before retrying. If it is missing, increase the delay between retries.You may not need this: the official SDKs retry on their own (see above). If you call the API directly, this helper waits and retries, using the server’s requested delay when you pass one. Treat this 503 as retryable; never retry billing or quota errors.
Python
import random
import time
def with_retries(call, is_retryable, retry_after=lambda err: None, attempts=5):
"""Run call(). On a retryable error, wait and try again, up to attempts tries.
retry_after(err) returns the server's requested wait in seconds, or None."""
for attempt in range(attempts):
try:
return call()
except Exception as err:
if attempt == attempts - 1 or not is_retryable(err):
raise
wait = retry_after(err)
if wait is None:
wait = min(60, 2 ** attempt) + random.random() # backoff + jitter
time.sleep(wait)
JavaScript
async function withRetries(call, isRetryable, retryAfter = () => null, attempts = 5) {
// retryAfter(err) returns the server's requested wait in seconds, or null.
for (let attempt = 0; ; attempt++) {
try {
return await call();
} catch (err) {
if (attempt >= attempts - 1 || !isRetryable(err)) throw err;
let wait = retryAfter(err);
if (wait == null) wait = Math.min(60, 2 ** attempt) + Math.random(); // backoff + jitter
await new Promise((resolve) => setTimeout(resolve, wait * 1000));
}
}
}
These AI APIs work with the official OpenAI library. Change the address and the key, pick a model from that provider’s list, and you’re running again. Each shows its live status, so you don’t switch to one that’s also down.
One key for many companies' models. Pick a model from a different company than the one that is down.
OpenRouter’s current model list · setup guide
Python
import os
from openai import OpenAI
client = OpenAI(
base_url="https://openrouter.ai/api/v1",
api_key=os.environ["OPENROUTER_API_KEY"],
)
reply = client.chat.completions.create(
model="PICK_FROM_LIST", # see OpenRouter's current model list
messages=[{"role": "user", "content": "Hello"}],
)
print(reply.choices[0].message.content)JavaScript
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const reply = await client.chat.completions.create({
model: "PICK_FROM_LIST", // see OpenRouter's current model list
messages: [{ role: "user", content: "Hello" }],
});
console.log(reply.choices[0].message.content);Groq’s current model list · setup guide
Python
import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.groq.com/openai/v1",
api_key=os.environ["GROQ_API_KEY"],
)
reply = client.chat.completions.create(
model="PICK_FROM_LIST", # see Groq's current model list
messages=[{"role": "user", "content": "Hello"}],
)
print(reply.choices[0].message.content)JavaScript
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.groq.com/openai/v1",
apiKey: process.env.GROQ_API_KEY,
});
const reply = await client.chat.completions.create({
model: "PICK_FROM_LIST", // see Groq's current model list
messages: [{ role: "user", content: "Hello" }],
});
console.log(reply.choices[0].message.content);Google's OpenAI-compatible mode for the Gemini API.
Gemini’s current model list · setup guide
Python
import os
from openai import OpenAI
client = OpenAI(
base_url="https://generativelanguage.googleapis.com/v1beta/openai/",
api_key=os.environ["GEMINI_API_KEY"],
)
reply = client.chat.completions.create(
model="PICK_FROM_LIST", # see Gemini's current model list
messages=[{"role": "user", "content": "Hello"}],
)
print(reply.choices[0].message.content)JavaScript
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://generativelanguage.googleapis.com/v1beta/openai/",
apiKey: process.env.GEMINI_API_KEY,
});
const reply = await client.chat.completions.create({
model: "PICK_FROM_LIST", // see Gemini's current model list
messages: [{ role: "user", content: "Hello" }],
});
console.log(reply.choices[0].message.content);Anthropic describes its OpenAI-compatible mode as mainly for testing; for long-term production use, its own SDK is recommended.
Claude’s current model list · setup guide
Python
import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.anthropic.com/v1/",
api_key=os.environ["ANTHROPIC_API_KEY"],
)
reply = client.chat.completions.create(
model="PICK_FROM_LIST", # see Claude's current model list
messages=[{"role": "user", "content": "Hello"}],
)
print(reply.choices[0].message.content)JavaScript
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.anthropic.com/v1/",
apiKey: process.env.ANTHROPIC_API_KEY,
});
const reply = await client.chat.completions.create({
model: "PICK_FROM_LIST", // see Claude's current model list
messages: [{ role: "user", content: "Hello" }],
});
console.log(reply.choices[0].message.content);Test a backup in development before routing production traffic. Pick a model from a different company than the one that’s down. Your code talks to the provider directly; OutageRadar never sees your key. For a full automatic fallback setup, see TokenMargin.
History logged by OutageRadar.
Full history: Is ChatGPT down?
Source: Error codes (OpenAI docs). Last checked against official docs: Nov 10, 2026. If the docs change, they win.