Claude experiences another outage affecting several models.
Anthropic's Claude AI is experiencing signs of overload once again. On Monday morning, users attempting to interact with Claude faced failed requests and a "529 Overloaded" error message. This message indicates a temporary server-side problem and directs users to Anthropic's status page. A screenshot shared with TNW shows the error appearing in Claude's interface.
Anthropic's own report on the incident is more comprehensive and cautious. At 05:06 UTC on August 24, the company announced it was investigating increased errors affecting requests to Claude Mythos 5, Claude Fable 5, Claude Opus 5, and Claude Opus 4.8. This incident also impacted claude.ai, the Claude API, Claude Code, and Claude Cowork.
At 05:27 UTC, Anthropic stated it had identified the issue and was working on a solution. Its latest update, posted at 06:42 UTC, indicates the company is still trying to resolve the problem. The status page currently does not provide an estimated recovery time or a technical explanation.
The distinction is important. The "529 Overloaded" error reflects what some users are experiencing, but Anthropic has not confirmed that overload is the primary cause of the broader issue. At this time, the public record notes increased errors across various models and products, but does not clarify the reason for their occurrence.
A busy August for Claude
This isn't the first issue Claude has faced this month. Anthropic's status history records a performance degradation incident impacting multiple models on August 18, followed by another issue involving Opus 5 and Haiku 4.5 on August 19. A separate disruption affecting several models was noted on August 20.
Independent reports have also covered the latest disruption. Frandroid noted that some users were unable to access Anthropic’s models and that the company had pinpointed the problem. Discussions on Hacker News and other forums show users experiencing the same 529 error while trying to utilize Claude.
These reports do not clarify the extent of the outage, but they illustrate how swiftly an infrastructure issue becomes apparent: through a failed chat request, a broken coding session, or a retry of an API call.
The implications of shared infrastructure
Claude is now a suite of models presented through consumer chat, developer tools, an API, and Cowork. When multiple platforms fail simultaneously, users do not face isolated incidents; instead, they encounter a single unavailable dependency.
This is a crucial aspect of the situation. Anthropic has identified the issue but has yet to specify if the problem lies within a shared service, model routing, capacity, or another component of the system. Any speculation about the cause remains unconfirmed at this point.
The situation is still being monitored for a resolution. Until Anthropic provides further details, the question isn't just when Claude will be operational again; it is whether users can rely on it without anticipating failures as part of the interface.
For developers, this distinction may prove significant in the future. During the outage, the immediate practical question is simpler: should they wait, retry, switch to another provider, or close their laptops and spend time with their pets?
Other articles
Claude experiences another outage affecting several models.
The Claude outage impacts Mythos 5, Fable 5, Opus 5, and Opus 4.8, disrupting Claude.ai, the API, Claude Code, and Cowork.
