Anthropic's Claude API has recovered from a major outage, but Claude.ai login problems continue affecting users. The incident highlights ongoing reliability challenges for AI platforms.
The Claude AI platform experienced another significant outage today, with the API recovering while Claude.ai login issues continue to affect users. According to the official status page, the Claude API has fully recovered as of 8:01 PT / 16:01 UTC, but problems persist with the web interface and authentication systems.
Current Status Breakdown
As of the latest update at 15:22 UTC, the platform shows a mixed reliability picture:
- Claude API (api.anthropic.com): Operational - fully recovered
- Claude.ai (Web Interface): Degraded - ongoing login issues
- Claude Code CLI Tool: Degraded - login broken, but logged-in users can still use it
- Claude Cowork Service: Operational
- Claude for Government: Operational
Incident Timeline
The disruption began earlier today at 14:53 UTC when elevated errors were first detected across Claude.ai, the API, and Claude Code. The incident follows a pattern of recent reliability challenges for the platform, with multiple outages recorded throughout April 2026.
Impact on Users
Claude Code users who are already logged in can continue using the tool, but new logins are currently broken. This creates a particularly frustrating situation where existing sessions remain functional while new users cannot access the service. The Claude.ai web interface is also experiencing degraded performance, primarily affecting login functionality.
Reliability Metrics
Looking at the broader picture, Claude's reliability metrics show room for improvement:
- 30-day uptime for Claude.ai: 89.48%
- 30-day uptime for Developer Console: 91.37%
- 30-day uptime for Claude API: 91.18%
- 30-day uptime for Claude Code: 90.86%
- 7-day uptime average: 93.01%
Global latency remains relatively low at an average of 78ms across 17 regions, with the fastest response times in Hyderabad (16ms) and Tokyo (22ms).
Recent Outage History
The current incident is part of a series of disruptions in recent weeks. Notable recent events include:
- April 7, 2026: Elevated errors on Claude.ai affecting login and voice mode
- April 6, 2026: Login issues for Claude.ai and Claude Code
- April 4, 2026: Elevated error rates for Sonnet 4.6 and Opus 4.6 models
- April 3, 2026: Issues with Claude Sonnet 4.6 requests
Technical Response
Anthropic's status page indicates that the root cause has been identified and fixes are being implemented. The company is working to mitigate ongoing errors for Claude AI while maintaining API functionality. This approach of prioritizing API recovery while addressing web interface issues suggests a staged recovery strategy.
Industry Context
These reliability challenges come at a time when AI platforms are under increasing pressure to deliver consistent, enterprise-grade performance. As more businesses integrate AI tools into their workflows, even brief outages can have significant productivity impacts.
The incident also highlights the complexity of maintaining large-scale AI infrastructure, where different components (web interfaces, APIs, CLI tools) can experience varying levels of disruption simultaneously.
Looking Ahead
While the API recovery is positive news for developers and businesses relying on Claude's backend services, the continued login issues for the web interface and Claude Code represent a significant user experience problem. Anthropic will need to address these authentication challenges quickly to restore full platform functionality.
For users affected by the outage, the recommendation is to use the API if possible, as it has fully recovered, or to wait for the login issues to be resolved before attempting to access Claude.ai or Claude Code.
The incident serves as a reminder of the ongoing reliability challenges in the AI infrastructure space, even for well-funded and technically sophisticated companies like Anthropic.
Comments
Please log in or register to join the discussion