# YOLOZU

Status: Released evaluation toolkit: v4.11.0.

Validate predictions.json artifacts and compare vision models on the same dataset and metric protocol, through CLI, Python and local MCP interfaces.

Stable prediction evaluation is separate from training, benchmark and research workflows. Synthetic examples verify the workflow, not model quality. The adaptive image service remains gated on qualification and activation; it is not a ready-to-use hosted detector.

Canonical page: [YOLOZU](https://www.toppymicros.com/yolozu/)

Primary sources:

- [Release v4.11.0](https://github.com/ToppyMicroServices/YOLOZU/releases/tag/v4.11.0)
- [Python package](https://pypi.org/project/yolozu/)

Agent resources:

- [Agent guide (Markdown)](https://www.toppymicros.com/yolozu/docs/agents.md)
- [Capabilities and qualification limits (JSON)](https://www.toppymicros.com/yolozu/docs/capabilities.json)
- [Documentation index](https://www.toppymicros.com/yolozu/docs/llms.txt)

Reviewed: 2026-10-11

Source guide: [Toppy AI resources](https://www.toppymicros.com/ai/)