--- title: "Is it Anthropic or Selan? status.selan.ai now tells you · Selan" description: "See whether Claude, OpenAI, Vertex AI, OpenRouter and Kimi are answering Selan's requests right now, beside what each provider says, with 30 days of their incidents and Selan's own status." url: https://selan.ai/blog/2026-11-13-status-page updated: 2026-11-13T16:10:13+02:00 --- [← All posts](https://selan.ai/blog) Blog # Is it Anthropic or Selan? status.selan.ai now tells you See whether Claude, OpenAI, Vertex AI, OpenRouter and Kimi are answering Selan's requests right now, beside what each provider says, with 30 days of their incidents and Selan's own status. 13 November 2026FeatureIn the docs: [Status page](https://selan.ai/status) When Claude Code stops with "API Error", you want to know whose fault it is: Anthropic's, your network's, or Selan's in between. The provider's status page is the obvious place to look, and it is often the last to know. It says Operational until somebody there writes an incident. On 13 November that gap was twelve minutes. From 09:46 UTC, Anthropic's sign-in service began refusing some of the requests made with Claude subscriptions, answering "credential validation failed". Claude Code retried each one, and some people saw API Error once the retries ran out. Anthropic posted its incident at 09:58 UTC. Selan's own traffic had shown the failures from the first minute, but nobody outside Selan could see them. ## Every provider, as your requests see it [status.selan.ai](https://status.selan.ai) measures each AI provider from the requests Selan sends it, and puts what the provider says about itself alongside. ![The top of status.selan.ai: Everything is running, then the AI providers list with Claude with a subscription, OpenRouter, OpenAI Codex and Kimi, each with the share of requests that succeeded, how soon answers started, 30 daily bars and what the provider's own page says](https://selan.ai/img/blog/2026-11-13-status-page/status-page.webp) Each row is one way Selan reaches a provider: Claude with a subscription and with an API key, OpenAI's Codex and its API, Google Vertex AI, OpenRouter and Kimi. A row shows: - the share of requests that succeeded in the last 15 minutes, and how soon answers started; - the last 30 days, a bar a day, from the provider's own incidents on the parts Selan uses; - what the provider's status page says right now, with a link to it. A failure counts against the provider only when the provider answered with it: an error from Anthropic, or a stream it broke off. Anything that goes wrong inside Selan counts against Selan, which has a row of its own under the providers, with 30 days of history. On a day like 13 November the page reads "Claude is having trouble", and under it "Selan itself is running normally". ## Open a row for the detail A row opens to the last 24 hours as Selan saw them, half an hour a bar, and the incidents the provider reported over 30 days, each linked to the provider's own page. ![Claude with a subscription opened on status.selan.ai: the last 24 hours as Selan saw them, the last 15 minutes, the incidents Anthropic reported over 30 days with their dates, and Partial outage as status.claude.com says now](https://selan.ai/img/blog/2026-11-13-status-page/claude-row.webp) Claude with a subscription, opened. Under the providers, the page lists open incidents in the providers' own words. It shows only those on the parts Selan uses, so an outage of a connector you never call stays off it. The numbers are totals across all of Selan's traffic, and no company, person or prompt is ever shown. The page updates every minute. It is also at [selan.ai/status](https://selan.ai/status), linked from every page's footer. **How it works today.** A post says what changed on its day; the docs say how Selan works now. [Status page](https://selan.ai/status)