# Tessary duration drift

duration_drift is one of Tessary's built-in classifiers. It watches each call site's latency against
its own past, covering turn duration and per-tool duration.

It catches the slowdowns that creep in without an outage: a model that got slower, a tool that
started dragging, a prompt change that added seconds to every turn. Comparisons run against the call
site's own history, both its recent normal and a reference pinned at the last deploy, so sudden
breaks and slow creep both register. A detection reads as a plain multiplicative shift, such as 1.4x
slower. Findings go to triage, where the ones ruled real issues open cases.

## Questions answered under this concept

- [Can duration_drift catch a slowdown that happens gradually, not all at once?](https://tessary.ai/answers/tessary-duration-drift/can-duration-drift-catch-a-gradual-slowdown)
- [Can duration_drift tell me which tool call made a turn slow?](https://tessary.ai/answers/tessary-duration-drift/can-duration-drift-tell-me-which-tool-is-slow)
- [Does duration_drift use a fixed latency threshold?](https://tessary.ai/answers/tessary-duration-drift/does-duration-drift-use-a-fixed-threshold)
- [Is duration_drift turned on by default?](https://tessary.ai/answers/tessary-duration-drift/is-duration-drift-turned-on-by-default)
- [What is Tessary's duration_drift classifier?](https://tessary.ai/answers/tessary-duration-drift/what-is-tessarys-duration-drift-classifier)
- [Why did my AI agent get slower even though nothing is throwing errors?](https://tessary.ai/answers/tessary-duration-drift/why-did-my-agent-get-slower-with-no-errors)

---

Source: https://tessary.ai/answers/tessary-duration-drift
All concepts: https://tessary.ai/answers
From Tessary, agent reliability for AI agents in production: https://tessary.ai