Status monitoring Alert setup guide

Get notified when AI services go down

Stop refreshing status pages manually. Set up automatic email, Discord, or Slack alerts for OpenAI, Anthropic, Cursor, and 77+ more AI providers — free, 2-minute setup.

Why you need status alerts

AI outages cost real time

When OpenAI or Anthropic goes down mid-deployment, debugging "why isn't this working?" wastes 20-40 minutes. An alert in the first 2 minutes saves the whole session.

Providers bury incident info

Most AI status pages are slow to update, hard to navigate, and scattered across different URLs. Prismix normalises them all into one dashboard.

Incidents are more common than you think

Major AI providers average 2-4 incidents per month. Even "minor degradation" can cut throughput by 50%+ or cause cryptic 500 errors in production.

Your team needs the same information

Webhook alerts let you push status changes to a shared Slack or Discord channel so your whole team knows — not just whoever happened to notice.

Alerts for 77+ AI providers

Star any of these services to receive alerts when their status changes:

See all 77 services →

2-minute setup

  1. 1

    Create a free account

    Free

    Sign in with email or GitHub at prismix.dev/sign-in. No credit card, no verification wait.

    Sign in →
  2. 2

    Star the AI services you care about

    Free · up to 5 services

    On the status dashboard, click the ☆ star next to each provider. Free accounts get up to 5 starred services.

    Go to dashboard →
  3. 3

    Configure alert preferences

    Free

    Go to Alerts and set your severity threshold: "Any degradation" (more sensitive) or "Major outage only" (less noise). Email alerts go to your account address automatically once you star a service.

    Alert settings →
🔔

Discord and Slack alerts — Pro

Pro adds webhook delivery to Discord, Slack, or any HMAC-signed endpoint. One Prismix Pro account can cover your whole team.

🎮

Discord webhook

Paste your webhook URL — alerts appear in a channel as formatted embeds with indicator color (red = outage, yellow = degraded, green = recovery).

💬

Slack webhook

Incoming webhook URL from Slack App settings. Same formatting, with service name and incident summary in the message.

🔗

Generic webhook (HMAC-signed)

Any HTTPS endpoint. POST with JSON payload + X-Prismix-Signature header so you can verify authenticity in your own pipeline.

Daily and weekly digest emails

Pro accounts also get a combined digest that batches status incidents, top AI news, and MCP server releases into one email — delivered at a time you choose, in your timezone. Less noise, same coverage.

Prismix Daily Digest — June 12, 2026

─────────────────────────────

🔴 OpenAI: API latency incident (started 14:23 UTC, resolved)

🟡 Anthropic: Degraded performance (Claude.ai) — ongoing

✅ Cursor, GitHub Copilot, Mistral — all operational

─────────────────────────────

📰 Top news: GPT-4.5 context window extended, Anthropic raises Series F...

FAQ

How quickly do alerts fire after an incident starts?
Status data refreshes every ~5 minutes via cron. Alert dispatch fires within seconds of a state change being detected. End-to-end latency is typically under 6 minutes.
Can I get an alert when a service recovers?
Yes — by default you get both the degradation/outage alert and the recovery notification when the indicator returns to operational.
How many services can I watch on the free plan?
Up to 5 starred services on free. Pro removes this cap and adds multiple email destinations, webhooks, quiet hours, and per-service threshold overrides.
Does it work with anonymous email subscriptions?
Yes — on any /service/[provider] page there is a subscribe button for anonymous email alerts. No account needed for basic outage notifications on a single service.
Can I filter alerts by severity?
Yes. The global threshold (Degraded and worse / Major outage only) filters on the free plan. Pro adds per-service overrides so you can ignore minor degradation on OpenAI but alert immediately on Anthropic.