# Output Drift

Source: https://customlabs.io/glossary/output-drift/
Updated: 2026-09-13

Cost & ops

# Output Drift

Also known as model drift, quality drift

Output drift is a gradual change in a live system's output, with no release behind it.

Causes include a provider adjusting a model, a retrieval corpus shifting, or edge cases the eval set never covered.

It is slower and less visible than a hard regression.

It needs its own threshold and a named owner, not the alerts that only catch a crash.

[The AI Release Path](https://customlabs.io/release/)

[← Back to the full glossary](https://customlabs.io/glossary/)

## Related terms

[Observability Observability captures traces of every prompt, retrieval, tool call, and response for debugging.](https://customlabs.io/glossary/observability/)[Canary Release A canary release exposes a change to a small, deliberately chosen slice of traffic first.](https://customlabs.io/glossary/canary-release/)

## More in Cost & ops

[Inference Cost Inference cost is what it costs to run a trained model on a request.](https://customlabs.io/glossary/inference-cost/)[Notebook-to-Production Notebook-to-production is the gap between a working prototype and a system that runs reliably.](https://customlabs.io/glossary/notebook-to-production/)[Model Deprecation Model deprecation is a provider retiring a model version on its own schedule.](https://customlabs.io/glossary/model-deprecation/)[Acceptance Rate Acceptance rate is the share of an agent's proposed changes a human reviewer approves.](https://customlabs.io/glossary/acceptance-rate/)