# Inspect Evals

> Community collection of ready-to-run evaluation tasks for the UK AISI Inspect AI framework.

- id: `inspect-evals`
- category: Evals & Observability (`evals-observability`)
- url: https://ukgovernmentbeis.github.io/inspect_evals/
- repo: https://github.com/UKGovernmentBEIS/inspect_evals
- docs: https://ukgovernmentbeis.github.io/inspect_evals/
- status: active
- pricing: open-source
- license: MIT
- tags: evals, benchmarks, inspect-ai, python, open-source
- auth: none
- access notes: pip install inspect-evals; requires Inspect AI. Model credentials depend on chosen tasks.
- related: Inspect (https://indexagentica.com/entries/inspect-ai/)
- sources: https://github.com/UKGovernmentBEIS/inspect_evals , https://github.com/UKGovernmentBEIS/inspect_evals/blob/main/LICENSE , https://pypi.org/project/inspect-evals/ , https://ukgovernmentbeis.github.io/inspect_evals/
- added: 2026-11-09
- updated: 2026-11-09
- last verified: 2026-11-09
- maintainer: UK AI Security Institute
- submitted by: agentica-curator
- page: https://indexagentica.com/entries/inspect-evals/
- json: https://indexagentica.com/api/entries/inspect-evals.json

MIT-licensed eval suite (UKGovernmentBEIS/inspect_evals; not archived). PyPI inspect-evals 0.28.0 and docs https://ukgovernmentbeis.github.io/inspect_evals/ verified 2026-11-09. Install alongside Inspect AI; individual benchmark datasets and any model APIs retain their own terms and costs. Established companion to Inspect, not marketed here as a brand-new product launch.
