API
Build on top of Prismix. Free, public, no key required. We cache aggressively so a thousand consumers don't fan out into a thousand status-page fetches.
Status snapshot
GET https://prismix.dev/api/v1/statuses
Live status of every tracked AI provider. 60-second KV cache. Response shape:
{
"fetchedAt": "2026-05-11T12:00:00.000Z",
"cached": true,
"services": [
{
"id": "anthropic",
"name": "Anthropic",
"indicator": "none", // none | minor | major | critical | unknown
"description": "All Systems Operational",
"reachable": true,
"activeIncidents": 0,
"uptime30dPct": 99.97,
"recentIncidents30d": 2,
"lastIncidentAt": "2026-05-08T14:23:00Z",
"recentIncidentBriefs": [
{ "id": "abc", "name": "Elevated latency on Claude API",
"impact": "minor", "createdAt": "...", "resolvedAt": "..." }
]
}
]
} Embeddable badges
GET https://prismix.dev/api/badge/{serviceId}.svg
Shields-style live SVG. 30-second cache. Drop into READMEs, docs, dashboards. Try it:
Markdown
[](https://prismix.dev/service/anthropic) Light theme
/api/badge/{id}.svg
Dark theme
/api/badge/{id}.svg?theme=dark
Service IDs
We currently track 48 AI providers across model APIs, dev tools, infrastructure, and consumer apps. See the full list at /status. Each service URL on the dashboard is the canonical id (e.g. /service/anthropic → anthropic).
Rate limits + fair use
- Responses are cached at the edge for 30–60 seconds. Polling more than once a minute returns the same payload.
- No auth required, no key required, no per-IP cap today. We reserve the right to add one if anyone abuses it.
- Attribution is appreciated, not required. A link back to
prismix.devif you embed badges is the nice thing. - Status data is sourced from each provider's own public feed. Prismix is not affiliated with any listed service.
What's next
Per-service detail endpoint with incident timeline (/api/v1/services/{id}), webhook subscriptions for state changes, and an OpenAPI spec are on the roadmap. Vote them up if you'd use them.