# openai/evals

Evals is a framework for evaluating LLMs and LLM systems, and an open-source registry of benchmarks.

Repository: https://github.com/openai/evals
Canonical: https://ross.abutalabs.com/products/evals
Language: Python
License: NOASSERTION
License Family: other
Last push: 2026-04-14T15:29:57+00:00

## Health v2 (maintenance only)
Score: 66/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 77, release rhythm 35, longevity 94
- inputs: {"age_days": 1318, "days_push": 141, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 19262, forks 3060 (observed 2026-08-28T04:11:27.539465+00:00)

## What it is
OpenAI Evals is a Python framework for evaluating large language models and LLM-based systems, with an open-source registry of benchmark evals. It supports running existing evals and writing custom ones, including private evals built on your own data.

## Use cases
- evaluate llm model quality across benchmarks
- compare how different model versions affect my use case
- build custom evals for my llm application
- run benchmark suite on openai models
- create private evals from my own data
- test llm system behavior before deploying

## When to choose
- you are building with OpenAI models and want standardized evaluation
- you need a registry of ready-made LLM benchmarks
- you want to write custom evals in Python for your LLM workflows

## When to avoid
- you need evaluation of non-OpenAI models without adaptation
- you want a GUI-only evaluation experience (the OpenAI Dashboard may suffice instead)
- you are not working with LLMs at all

## Facets
- artifact type: framework
- maturity: active
- function: testing, benchmarking, llm-inference, prompt-engineering
- domain: large-language-models, machine-learning, developer-tools, artificial-intelligence
- platform: python, cli, cross-platform
- tags: llm-evaluation, benchmarks, openai, eval-registry, model-testing

## Member repositories
- openai/evals (main) score 66

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:27.539465+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:00:23.878559+00:00, confidence not recorded.
  - readme: https://github.com/openai/evals (fetched 2026-08-28T04:11:27.539465+00:00, sha c702dda2917e)
  - registry_pypi: https://pypi.org/pypi/evals/json (fetched 2026-08-29T07:58:40.147547+00:00, sha 3263beae54c1)
- Data as of 2026-08-30T08:39:29.467469+00:00.