Skip to main content
SLO MONITORING

SLO
Monitoring

Set targets for uptime, latency, and error rate. Get alerted when things go off track.

See It in Action

Watch how SLOs track your metrics and fire alerts when targets are breached

SLO Monitoring

Uptime

99.97%

Target: 99.95%

HEALTHY

P95 Latency

487ms

Target: <2,000ms

HEALTHY

Error Rate

0.05%

Target: <0.10%

HEALTHY

Uptime Trend

100%99.9599.90-60m-30mnowTarget

Error Rate Trend

0.20%0.10%0%-60m-30mnowTarget

99.95%

Uptime Tracking

Configurable target threshold

P50/P95/P99

Latency Percentiles

Full distribution visibility

<1%

Error Rate Monitoring

Track failures over time

Custom

Configurable Targets

Set your own thresholds

What You Can Track

Three metrics that tell you if your LLM integration is healthy or not.

Uptime

The percentage of requests that succeed over time. Set a target like 99.95% and know immediately when you drop below it.

Latency

P50, P95, and P99 response times. See how fast your requests are for most users, and catch the slow outliers too.

Error Rate

The percentage of requests that fail. Whether it’s provider errors or rate limits, you’ll see the trend and get alerted.

How Alerts Work

You set the target. Costbase watches it. When something breaks, you hear about it right away.

01

Set a Target

Pick a metric and define what "good" looks like. For example, 99.95% uptime or P99 latency under 2 seconds.

02

Continuous Tracking

Costbase measures every request against your target. No sampling, no delays. Every single call counts.

03

Violation Detected

When a metric drops below your target, Costbase flags it immediately. No waiting for a dashboard refresh.

04

Alert Sent

You get notified on Slack, Discord, email, or via webhook. Your team knows before users start complaining.

Who This Is For

If you care about reliability, SLO monitoring gives you the numbers to back it up.

Platform Teams

Track provider reliability across every model and endpoint. Know which providers are meeting their promises and which aren’t.

SRE Teams

Set SLO-based alerts that match how you already think about reliability. No more guessing if things are "mostly fine."

Engineering Leads

Get latency visibility without digging through logs. P50, P95, P99 — all in one place, updated in real time.

Product Teams

Understand the experience your users are actually getting. If response times are creeping up, you’ll know before they do.

Stop Guessing. Start Measuring.

Set up SLO monitoring in minutes. Know exactly how your LLM integration is performing.

Frequently Asked Questions