We built a free status monitor for 77 AI APIs. Here's what 6 weeks of data taught us.

Every AI developer has been here: your app is throwing 503s, users are pinging you, and you have 12 browser tabs open — OpenAI status page, Anthropic status page, the GitHub Copilot health page, three different Discord servers — trying to figure out is this me or is it them?

That's the problem we set out to solve. Prismix aggregates status from 77 AI services in one place. Six weeks of running it in production taught us some things that might save you time.

The problem is worse than you think

AI APIs don't fail like traditional infrastructure. They fail in weird, partial ways:

• Degraded performance that passes your health checks but makes your product feel broken
• Regional outages — OpenAI US-East is down while EU is fine, so half your users are affected
• Silent rate-limit cascades — the API returns 429s but their status page says "operational" for another 20 minutes
• Incident lag — providers often post status updates 10–30 minutes after engineers are already aware

The official status pages are optimistic by design. They're customer-facing communications tools, not real-time engineering dashboards.

What 77 status pages look like in aggregate

When you watch 77 AI services simultaneously, patterns emerge fast.

OpenAI is the most-watched service. The pattern is almost always: investigating → identified → monitoring → resolved, typically in 45–90 minutes. The investigating phase is where most developers panic — it looks bad but usually resolves without action on your end.

Anthropic runs noticeably clean compared to its API usage growth. Incidents are rarer and shorter.

The long tail is interesting. Services like Replicate, Runway, ElevenLabs, and Suno have incident patterns that don't correlate with OpenAI at all — genuine independent failure domains, useful for redundancy.

The "silent degradation" problem is real. Multiple times we've seen a service show "operational" while our uptime probe was timing out. This is why Prismix shows a latency sparkline per service — the status page catches announced incidents, the probe catches real ones.

What Prismix built (and why it's free)

Prismix pulls from official status pages and adds things the individual pages don't have:

Per-service latency probes — 24-hour sparklines showing actual response times, not just announced incidents.

Cross-service incident timeline — /incidents shows everything across all 77 services in one scrollable feed. Useful for postmortems.

Public REST API — GET /api/v1/statuses returns current status as JSON. No auth, no rate limit, CORS open. Free forever.

MCP server — ask Claude "Is OpenAI down right now?" and get a live answer. One line in claude_desktop_config.json.

It's free because it runs on Cloudflare's free tier (Workers + KV). The Pro tier ($10/mo) adds email/webhook alerts, but the core dashboard stays free.

The technical part

The stack is Astro 5 SSR + Cloudflare Workers + KV. KV's free tier gives 100,000 reads per day but only 1,000 writes. The cron job refreshing status runs every 5 minutes, so every write is conditional — only write if the content actually changed. That dropped writes from ~8,400/day to ~600/day. Monitoring infrastructure has to be cheap to run, otherwise the incentive to keep it free disappears.

What we don't know yet

Six weeks in, some things we're genuinely uncertain about:

• Which services are missing? The list covers LLM APIs and popular AI tools but we've probably missed something in your stack.
• Is the latency probe useful? It tells you "this service is slow right now" but not "slow compared to what" — no historical baseline yet.
• What would make you actually use this every day? A Slack bot? A PagerDuty integration? Something in your terminal?

If any of that resonates, drop a comment. Live at prismix.dev.

Also at Prismix: an MCP server directory with 500+ servers and a curated AI news feed — but the status monitoring is the part we're most curious to hear about.