Is Claude Code down, or is it your setup?
The Claude Code and Claude API rows from status.claude.com are below. During an outage Claude Code shows "API Error: 529" and retries on its own. If both rows are green, run claude doctor and /login. If Anthropic is down, switch model with /model or run Claude Code on another provider.
Live Claude Code status
Checked
All systems operational
- Claude CodeOperational
- Claude API (api.anthropic.com)Operational
- claude.aiOperational
- Last incident: Elevated errors on platform.claude.comresolved Oct 13
Source: status.claude.com. Read on our server about every 2 minutes.
Outage or local problem?
| What you see | Likely cause |
|---|---|
API Error: 529 with overloaded_error | Anthropic is over capacity. Check the Claude API row above |
API Error: 500 with api_error | An error at Anthropic. Retry, and check for an incident |
| A message that your usage limit is reached | Your Pro or Max window ran out. See Claude Code usage limits |
401, or a prompt to run /login | An expired session or a wrong key. Run /login |
claude: command not found | An install or PATH problem. Reinstall, then run claude doctor |
| Requests hang, then time out | Network, proxy or VPN. Try another network, or check HTTPS_PROXY |
Get Claude Code working again
- Run
claude doctorto check the installation, andclaude updateto get the latest version. - If the status page is green but requests fail with 401, run
/loginagain. - If an incident names one model, type
/modeland pick another one: Opus, Sonnet or Haiku. - If the whole Claude API is down, run Claude Code against another provider with the settings below.
Run Claude Code somewhere else
| Route | Settings |
|---|---|
| Amazon Bedrock | CLAUDE_CODE_USE_BEDROCK=1, AWS_REGION, and AWS credentials with access to the Claude models |
| Google Vertex AI | CLAUDE_CODE_USE_VERTEX=1, CLOUD_ML_REGION and ANTHROPIC_VERTEX_PROJECT_ID |
| Standard Compute | Run standardcompute launch claude, or set ANTHROPIC_BASE_URL=https://api.stdcmpt.com (no /v1), your key in ANTHROPIC_AUTH_TOKEN and ANTHROPIC_MODEL=anthropic/claude-sonnet-5.5 |
| A non-Claude model | On a gateway, type the ID: /model openai/gpt-6-astra or /model z-ai/glm-5.3. The picker only lists Claude models |
Bedrock and Vertex AI need your own cloud account with Claude enabled. Install the Standard Compute launcher with curl -fsSL https://standardcompute.com/cli/install.sh | sh.
What Standard Compute does when Anthropic fails
- Claude Code's requests go through OpenRouter to Anthropic's API or Google Vertex AI first, depending on the key. If that provider fails, OpenRouter can move the request to another provider of the same Claude model, such as Amazon Bedrock.
- Each failed call is retried twice on a 429, a 5xx or a network error, after 0.5 and then 1 second, or after the provider's Retry-After up to 8 seconds. If every attempt fails, your agent gets HTTP 503 with "The model is temporarily unavailable. Please retry shortly."
- A specific model ID always stays that model; failover only changes the provider. Only the auto-router, the model
standardcomputewith no model selection, can switch to a different model. Claude Code's short names (sonnet, opus, haiku) go to the auto-router; set the full ID to pin.
The catch
If Claude fails on every provider, Standard Compute cannot serve Claude either, and /model with another vendor's ID is what keeps you working. A Claude Pro or Max subscription stays with Anthropic's own login; Bedrock, Vertex AI and gateways bill separately.
Where Standard Compute fits
Standard Compute is an OpenAI- and Anthropic-compatible LLM gateway for coding agents at https://api.stdcmpt.com/v1, with monthly plans from $19 (USD) and a fixed compute budget. One key serves Claude, GPT, Gemini, DeepSeek and GLM models, so moving off a struggling model during an outage takes one changed model ID.
FAQ
Why is Claude Code not working when the status page is green?
Most often it is local: an expired login (run /login), an old version (run claude update), or a network or proxy problem. Run claude doctor first. A short capacity spike can also return 529 before Anthropic posts an incident, so retry after a minute.
Does Claude Code work while claude.ai is down?
They are separate rows on status.claude.com. Claude Code depends on the Claude API and on its own Claude Code row, so check those two; a claude.ai incident on its own does not necessarily affect it.
Can I keep using Claude Code during an Anthropic outage?
Yes, on another provider: Amazon Bedrock with CLAUDE_CODE_USE_BEDROCK=1 or Google Vertex AI with CLAUDE_CODE_USE_VERTEX=1 if you have those accounts, or a gateway through ANTHROPIC_BASE_URL. On Standard Compute a pinned Claude model fails over between providers of that model, and any other catalog model is one /model command away.
Can I use my Claude Pro or Max subscription through a gateway?
No. A gateway such as Standard Compute bills its own plan, from $19 a month (USD) with a fixed compute budget. Your Anthropic subscription keeps working with Anthropic's own login.