Anthropic 529 overloaded_error: what it means and how to fix it

529 overloaded_error: Overloaded

How to fix it

1. Retry with exponential backoff and jitter (start around 1 to 2 seconds, cap around 60). The SDKs do two retries by default; raise max_retries for jobs that must finish.
2. For work that can wait, send it through the Message Batches API, which queues it instead of failing.
3. For user-facing paths, fall back to another model or provider after a few failed retries.
4. If you also see 429s after a sudden traffic jump, ramp traffic up gradually: Anthropic applies acceleration limits to sharp increases.

Is it down right now?

Check our live status page or the official one.

Retry it properly

import anthropic

# The SDK retries 429, 5xx and connection errors with exponential backoff and honors retry-after.
client = anthropic.Anthropic(max_retries=6)

try:
    msg = client.messages.create(model="claude-sonnet-5-5", max_tokens=1024,
                                 messages=[{"role": "user", "content": "Hello"}])
except anthropic.RateLimitError as e:      # 429 after all retries
    print("still rate limited", e.response.headers.get("retry-after"))
except anthropic.APIStatusError as e:      # 529 overloaded, 500, 504...
    print(e.status_code, e.response.headers.get("request-id"))

The official SDKs already retry connection errors, 429 and 5xx twice with exponential backoff and honor retry-after; raise max_retries for batch jobs.

Same kind of error at other providers

• OpenAI 500 server_error
• OpenAI 503 server_is_overloaded
• Google Gemini API 503 UNAVAILABLE
• Google Gemini API 500 INTERNAL

Other Anthropic (Claude) errors

• 429rate_limit_errorretry
• 400invalid_request_errorfix
• 400invalid_request_error: This model does not support assistant message prefill. The conversatiofix
• 400invalid_request_error: "thinking.type.enabled" is not supported for this model. Use "thinkingfix
• 400invalid_request_error: "thinking.type.disabled" is not supported for this model. Use "thinkinfix
• 400invalid_request_error: `thinking` or `redacted_thinking` blocks in the latest assistant messafix
• 400invalid_request_error: tool_choice: type "tool" and "any" are not supported for this model.fix
• 401authentication_errorfix
• 402billing_errorfix
• 403permission_errorfix
• 404not_found_error: The requested resource could not be found.fix
• 413request_too_largefix
• 500api_errorretry
• 504timeout_errorretry

Source

• Anthropic (Claude) error documentation: platform.claude.com/docs/en/api/errors (checked Oct 24, 2026)

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