The Pulse
Anthropic Resolves Errors Across Five Claude Models
Anthropic reported elevated errors affecting five Claude models and four services on October 25, with impact recorded from 06:05 to 06:35 UTC. A separate incident later that day affected Claude Fable 5.1, according to a second status notice
AI.info Team · Written with AI · fact-checked ·
Anthropic resolved an incident involving elevated errors across five Claude models on October 25. The company’s status record says the impact ran from 06:05 to 06:35 UTC and affected Claude.ai, the Claude API, Claude Code and Claude Cowork. The notice does not say that requests began failing. Anthropic later logged a separate incident involving Claude Fable 5.1 alone.
Five models, four Claude services
Anthropic opened its investigation at 06:14 UTC, naming Claude Mythos 5.1, Claude Fable 5.1, Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 5.5. The update said:
“We are investigating elevated errors on requests to Claude Mythos 5.1, Claude Fable 5.1, Claude Opus 5.5, Claude Sonnet 5.5, and Claude Haiku 5.5.”
That status update identified the five models under investigation; the incident record separately lists Claude.ai, the API, Claude Code and Claude Cowork as affected services. The recorded impact began at 06:05 UTC, before the investigation update appeared.
At 06:27 UTC, Anthropic said it had identified the cause of elevated errors on requests to Sonnet 5.5 and was working on a fix. At 06:35 UTC, it reported that a fix had been deployed and that it was monitoring recovery. The record marks the incident resolved at 07:53 UTC, while placing the impact window between 06:05 and 06:35 UTC. It does not describe the cause or explain the gap between the end of the impact window and the resolution notice.
A second incident hit Fable 5.1
Anthropic logged another elevated-error incident later on October 25, this time naming Claude Fable 5.1. The status page places its impact from 08:08 to 08:29 UTC, again across Claude.ai, the API, Claude Code and Claude Cowork. At 08:17 UTC, Anthropic posted that it was investigating elevated errors and had identified the cause, and that it was working on a fix. The company reported a deployed fix and monitoring at 08:30 UTC, then marked the incident resolved at 08:49 UTC.
The two records are distinct: the first names five models and gives a 30-minute impact window, while the later report names Fable 5.1 and records a 21-minute window. Neither notice says whether the incidents were related. The records also provide no error rates, count of affected customers or description of the underlying technical fault.
Other recent model-error reports
The October 25 incidents followed several model-error reports in the status history on October 24. One notice names Claude Fable 5 and Fable 5.1, with impact recorded from 05:00 to 05:35 UTC. Another names Claude Opus 5.5 and records impact from 00:23 to 00:34 UTC. A separate report names Fable 5.1, Sonnet 5.5 and Haiku 5.5, with impact from 23:28 to 23:59 UTC on October 23. Those records describe separate incidents and do not say that the events shared a cause.
Anthropic’s status page currently lists Claude.ai, Claude Console, the Claude API, Claude Code, Claude Cowork and Claude for Government as operational. Both October 25 incident notices are marked resolved.