# Claude Status — Full Documentation > Unofficial community-built status dashboard for Claude AI services. > Website: https://claudestatus.com ## Overview Claude Status is a third-party status monitoring dashboard that tracks the health and performance of Anthropic's Claude services. It provides enhanced monitoring capabilities beyond the official status page, including global latency measurements, community-driven issue reporting, and historical reliability insights. ## Monitored Services - **claude.ai** — Web interface for Claude - **platform.claude.com** — Developer console (formerly console.anthropic.com) - **Claude API (api.anthropic.com)** — Backend API for developers - **Claude Code** — CLI tool for developers - **Claude for Government** — Government cloud deployment ## Features ### Real-time Status Component status (Operational, Degraded, Outage) fetched from Anthropic's public Statuspage API every 2 minutes. ### Uptime History 30-day uptime percentage calculated from incident duration data. Shows daily breakdown with color-coded bars (green = no issues, amber = degraded, red = outage). ### Reliability Insights Rolling uptime percentage with 7D/14D/30D toggle. SVG chart generated from real daily uptime data. ### Global Latency HTTP latency to claude.ai measured from 17 countries via check-host.net. Multiple nodes per country for redundancy. Data refreshed via Vercel Cron every 5 minutes. ### Community Monitoring Users can report outages and latency issues. Reports stored in Cloudflare KV with 3-day retention, 2000 entry cap. 24-hour trend charts available. ### Subscriptions - RSS feed at /rss.xml - Email subscription via Resend ## API Endpoints - GET / — Main dashboard (SSR) - GET /rss.xml — RSS feed - GET /api/latency — Cached latency data - POST /api/latency — Smart refresh with rate limiting - GET /api/report — Community reports + trend data - POST /api/report — Submit a report - POST /api/subscribe — Email subscription ## Static Pages - /about — Site purpose, who runs it, editorial principles, what is monitored - /methodology — Exact math, data sources, and trade-offs behind every dashboard number - /disclosure — Advertising, affiliate, sponsored-content, and data-sources policy - /contact — Maintainer inbox and topic guide - /privacy — Privacy policy covering data collection, Google Analytics, Google AdSense cookies, third-party services, GDPR/CCPA/PIPEDA/LGPD rights, children's privacy - /terms — Terms of service ## Blog Long-form posts on Claude reliability, status-page reading, and SRE practice. Each post is 1,000–1,500 words. - /blog/30-day-uptime-math — How we calculate 30-day uptime, and why our number is lower than the marketing figure (interval-merge downtime math, worked example, AWS/GCP/OpenAI comparison). - /blog/claude-529-overloaded-playbook — Reading Claude's 529 Overloaded errors: triage playbook, retry vs fail-fast, backoff math, correlating bursts with the public uptime feed. - /blog/partial-outage-taxonomy — What 'partial outage' actually means at Anthropic — a guided tour of the four Statuspage impact levels with examples. - /blog/17-country-latency-baseline — A 30-day latency baseline to claude.ai from 17 countries with regional anomalies. - /blog/claude-vs-openai-vs-gemini-status-pages — Side-by-side reading of all three frontier-model providers' public status pages. - /blog/community-report-sidebar — How to read the anonymous, rate-limited, three-day rolling community report sidebar during an incident. - /blog/cache-strategy-no-lag — The two-layer caching choice that keeps every page render under 100ms even when upstream Statuspage is hammered. - /blog/quiet-weekend-claude — What the data looks like when nothing is wrong, and why that baseline matters more than incident days. - /blog/reading-uptime-bars — Visual literacy guide for the 30-day uptime bars; the partial-day rule and what each shade means. - /blog/rss-for-status-pages — Why /rss.xml is built for machines, not humans — wiring incidents into Slack, PagerDuty, or your own dashboard. - /blog/claude-code-stability — Observations on Claude Code's stability curve since it joined the public Statuspage. - /blog/reading-statuspage-postmortems — How to read each section of a Statuspage postmortem and what audience each is written for. - /blog/best-of-three-vs-median — Picking the right summary statistic for global latency probes when you sample three nodes per country. - /blog/80-character-community-reports — Why 80 characters is the right cap for a community report (moderation, spam, readability). - /blog/incident-communication-templates — Six templates for telling your users when Claude is down, one per incident stage. - /blog/claude-api-error-codes — Working reference for every HTTP status code the Anthropic API returns and the right client reaction. - /blog/multi-provider-llm-failover — Patterns for falling back from Claude to GPT or Gemini in production, and the cost of resilience. - /blog/what-999-uptime-buys — Converting uptime percentages into real time, from 99% to 99.999% per month/quarter/year. - /blog/claudeai-loads-but-chat-doesnt — Debugging the gap between front-door reachability and inference availability. - /blog/trust-center-vs-statuspage — When to read Anthropic's Trust Center vs the public Statuspage during an incident. ## Languages & Internationalization Available in 9 languages with native, locale-specific copy — each rewritten for that market's real search terms (e.g. de "Störung", fr "en panne", ja "使えない/障害", zh "崩了/无法访问"), not literal translation. English is served at the site root; every other locale lives under a path prefix. The homepage and the four status pages (/claude-error, /claude-code, /api-status, /claude-cowork) are localized; each carries hreflang alternates (+ x-default), a language switcher, locale-aware internal links, and entries in the generated sitemap.xml. - English — / (default) - 日本語 (Japanese) — /ja/ - 简体中文 (Simplified Chinese) — /zh/ - 繁體中文 (Traditional Chinese) — /zh-tw/ - 한국어 (Korean) — /ko/ - Español (Spanish) — /es/ - Português (Brasil) — /pt/ - Deutsch (German) — /de/ - Français (French) — /fr/ ## Technology Stack - Astro 6 with SSR - Tailwind CSS 4 - Vercel (hosting + cron) - Cloudflare KV (data storage) - Resend (email subscriptions) - Google Analytics 4 (visitor analytics) - Google AdSense (advertising — see /privacy and /disclosure for full cookie policy) ## Disclaimer This is NOT an official Anthropic product. Not affiliated with, endorsed by, or connected to Anthropic. Data sourced from public APIs. For official status information, visit https://status.claude.com.