Anthropic’s Claude Hits Second Outage in Three Days — Is Claude Down
Claude is claude down again after a second major outage in three days on April 20, with access failures hitting Claude Chat, Claude Code, the API, and the mobile app starting at 14:53 UTC. Users saw “Service is temporarily busy” and “Page not found” messages while the disruption spread across the web interface and coding tools.
Downdetector complaints peaked at 6,500 users, which gives the outage a size that goes well beyond a routine hiccup. Anthropic acknowledged a Partial Outage and said it was seeing “Elevated errors on Claude.ai, API, Claude Code.”
Claude.ai and API recovery
Anthropic identified authentication issues as the primary problem, and it said the Claude API had fully recovered as of 8:01 PT / 16:01 UTC. By 15:20 UTC, the company said the API was back but access problems for the main interface were still ongoing.
That left Claude.ai users in a different position from API users. The service had stabilized for login success rates on Claude.ai, including via Claude Code, but the main web experience was still not fully back when the company last updated its status.
Three days, two outages
This was the second major Claude outage in three days, following a similar disruption earlier in April that left both conversational and coding tools inoperational for a substantial period. The repetition is the friction point here, because the outages hit multiple Claude products at once while enterprise adoption is rising.
For affected users, the practical takeaway is simple: Claude’s web app, coding tools, and API did not fail in isolation, and a return for one surface did not mean every surface was usable yet. The remaining open question is whether the repeated authentication failures will lead to any public explanation of the fix before the next incident appears.
