# braintrustdata/autoevals

AutoEvals is a tool for quickly and easily evaluating AI model outputs using best practices.

Repository: https://github.com/braintrustdata/autoevals
Canonical: https://ross.abutalabs.com/products/autoevals
Language: Python
License: MIT
License Family: permissive
Last push: 2026-07-29T20:54:43+00:00

## Health v2 (maintenance only)
Score: 85/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 95, release rhythm 75, longevity 82
- inputs: {"age_days": 1149, "days_push": 35, "days_rel": 85, "gap_med": 33.5, "n_releases_24m": 3}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1009, forks 78 (observed 2026-08-28T04:03:12.826211+00:00)

## What it is
Autoevals is a Python and TypeScript library for automatically evaluating AI model outputs using LLM-as-a-judge, heuristic, and statistical methods. It offers prebuilt evaluators like Factuality plus support for custom model-graded prompts.

## Use cases
- evaluate llm outputs automatically
- score ai model responses for factuality
- run llm-as-a-judge evaluations
- compute bleu or levenshtein scores for generated text
- build custom model-graded eval prompts
- test chatbot answer quality

## When to choose
- you need ready-made evaluators for subjective LLM tasks like factuality or safety
- you want a lightweight library usable from Python or TypeScript
- you want to tweak or debug evaluation prompts on individual examples

## When to avoid
- you need a full experiment-tracking or observability platform rather than just evaluators
- you cannot call an LLM API, since model-graded evals require one
- you only need deterministic string metrics with no LLM judging

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, llm-inference, testing, benchmarking, nlp
- domain: large-language-models, machine-learning, developer-tools, artificial-intelligence
- platform: python, cross-platform
- tags: llm-evaluation, llm-as-a-judge, evals, ai-quality, braintrust, nodejs

## Member repositories
- braintrustdata/autoevals (main) score 85

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:12.826211+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 2026-08-30T07:12:01.731246+00:00, confidence not recorded.
  - readme: https://github.com/braintrustdata/autoevals (fetched 2026-08-28T04:03:12.826211+00:00, sha 4161309c535f)
  - registry_pypi: https://pypi.org/pypi/autoevals/json (fetched 2026-08-29T13:12:09.055386+00:00, sha db78bd73b918)
- Data as of 2026-08-30T08:39:29.467469+00:00.