Claude

"Spanner temporarily unavailable"

Last checked

"Spanner temporarily unavailable" is a transient backend error on Anthropic's side, not something in your config or code. It often coincides with a broader service incident. Check the Anthropic status page to rule one out, wait about a minute, and retry. If it keeps recurring, switch to a different model such as Sonnet or Opus.

Likely cause

A transient backend error on Anthropic's side. Nothing in your config or code caused it; it often coincides with broader service incidents.

The fix

  1. 1 Check the Anthropic status page to rule out an active incident.
  2. 2 Wait about a minute and retry.
  3. 3 If it persists, switch to a different model (e.g. Sonnet to Opus).

Still failing?

  • Open status.claude.com and check the component you actually use (claude.ai, Claude Console, Claude API, Claude Code or Claude Cowork). Subscribe to the incident so you know when it clears.
  • On the API, let the official SDK retry instead of writing your own loop. It retries 5xx errors twice by default with exponential backoff; raise max_retries (maxRetries in TypeScript) for batch jobs.
  • Save the request-id response header (also returned as request_id in the error body) and include it if you contact support about repeated failures.
  • If the error has turned into 529 overloaded_error, that is documented API-wide high traffic, not a problem with your account. Back off and retry later.

Related errors

Hit a different error?

Paste any agent error and get the cause and fix in seconds.

Open the decoder

Frequently asked questions

Did I cause this with my config or prompt?

No. Anthropic does not document this message publicly, but it comes back from Anthropic's servers and clears without any change on your side. Nothing in your configuration or prompt triggers it.

Should I keep retrying immediately?

Wait about a minute first. Hammering retries during an incident rarely helps and can compound rate-limit pressure. Check the Anthropic status page while you wait.

It keeps happening. What else can I do?

If a single model keeps failing, switch to a different one (for example Sonnet to Opus) so a localized backend issue doesn't block all your requests.

Is it the same as a 529 overloaded error?

Not exactly. 529 overloaded_error is Anthropic's documented code for API-wide high traffic. "Spanner temporarily unavailable" is not in the documented error list, so handle it like a 500 api_error: retry with exponential backoff, and contact support with the request ID if it persists.

Stop firefighting agent errors

Decoding errors one at a time is the manual version of what BetterClaw automates. Run your agents on a no-code AI agent platform with managed models, retries and config validation built in.

Deploy Your Agent — Start FreeBook a Demo

Free plan available · Pro $49/mo · BYOK · 7-day money-back guarantee