Claude is in a major outage across API, Console, Claude Code, and claude.ai

Anthropic is reporting a major Claude outage affecting the services builders actually use: the Claude API, Claude Console, Claude Code, Claude Cowork, and claude.ai. This matters now because teams running Claude-backed products, coding agents, support workflows, or internal tools may see elevated errors until Anthropic's fix lands.

This is an official status-page incident, not social-media chatter. Anthropic opened the incident at 14:19 UTC on June 23, 2026, then updated it at 14:25 UTC to say the issue had been identified and a fix was being implemented.

What Anthropic says is affected

The incident is titled “Elevated error rate across multiple models.” Anthropic marks the impact as critical on the incident record, while the public Claude status summary shows a Partial System Outage.

Affected components are listed as:

claude.ai — major outage;
Claude Console at platform.claude.com — major outage;
Claude API at api.anthropic.com — major outage;
Claude Code — major outage;
Claude Cowork — major outage.

The latest status update available during this publish said: “The issue has been identified and a fix is being implemented.” Anthropic has not yet published a root cause or a resolved timestamp.

Builder impact

If your product depends on Claude, treat this as an active reliability incident, not a minor UI glitch.

Practical steps for engineering teams:

route critical traffic to a fallback model/provider if your product supports failover;
pause long-running Claude Code or agent jobs that are not idempotent;
increase retry backoff rather than hammering the API;
show degraded-mode messaging to users instead of generic failures;
avoid pushing Claude-dependent migrations, batch jobs, or customer demos until the status page moves to monitoring or resolved.

For teams using Claude Code as part of the development loop, expect interruptions in code-generation, repo-editing, and review flows. For SaaS products using the Claude API, the risk is customer-visible errors or slow recovery depending on how aggressively the app retries failed calls.

Caveats

Anthropic has only confirmed elevated error rates across multiple models and the affected components. It has not said which model families are most affected, how many requests are failing, whether data plane or control plane systems are involved, or when full recovery is expected.

There is also a separate OpenAI status issue at the time of writing: OpenAI is investigating elevated errors for ChatGPT file uploads and downloads. That is worth watching, but the Claude incident is the more immediate builder outage because Anthropic lists the Claude API and Claude Code as major-outage components.