Component of Claude

Is Claude Claude API (api.anthropic.com) down?

The status Claude publishes for Claude API (api.anthropic.com), on its own.

Operational

Official timeline

Active incidents naming this component

incidentElevated errors for multiple models
monitoring
Timeline →
incidentIntermittent error spikes for Claude Mythos 5.1 and Claude Fable 5.1
monitoring
Timeline →
incidentElevated errors for Claude Mythos 5.1 and Claude Fable 5.1
monitoring
Timeline →
incidentElevated latency on the Claude API for some users
identified
Timeline →
incidentElevated errors for multiple models
monitoring
Timeline →
incidentElevated errors for Claude Sonnet 5
monitoring
Timeline →
incidentElevated errors for Claude Sonnet 5
investigating
Timeline →

Collected history

Past incidents

IgniteStatus has not collected an incident naming Claude API (api.anthropic.com) yet. That is the limit of this record, not a claim about Claude’s history.

Data confidence

Source & freshness

Source: Claude Status ↗
Collected: Sep 22, 2026, 2:12 AM UTC
Provider updated: Sep 22, 2026, 2:11 AM UTC

Claude publishes this component separately, so its status can differ from the overall one. Operational means no issue is currently reported for it, not proof that every user or region is unaffected.

While this is open

What should I check next?

Read the incident timeline. The provider updates say what is affected and what they have already ruled out.
Match the affected components against the calls your system actually makes before assuming this incident is yours.
Hold off on rolling back your own deploy. A rollback during a provider incident removes a variable you will want later.
Queue or fail over the work if you can. Aggressive retries against a degraded provider usually make the backlog worse.
Watch the timeline rather than polling the API. It updates as the provider posts, including the resolution.