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

AutoEval: Autonomous Evaluation of Generalist Robot Manipulation Policies in the Real World | CoRL 2025

Automatic model evaluation (AutoEval) in CVPR'21&TPAMI'22

Trustworthy Evaluation of Robotic Manipulation: A New Benchmark and AutoEval Methods

A framework to generate automated evals for your agentic application.

dify-autoeval

scikit-autoeval: automatic evaluation for ML models

An autoevals-inspired toolkit for scoring AI model responses in Go.

AutoEval AI – Automatic Exam Paper Evaluation System

A minimal, vitest-native evals library for LLM applications. Built-in metrics, G-Eval LLM-as-judge, autoevals integration, and pretty console output.

AutoEval

LLM_AutoEval

TypeScript-first LLM evaluation library built on Autoevals with Vitest integration

AutoEval processes forms and data

Biocentral Hub: Autoeval Service (plm leaderboard)

Narrative summary coverage of underlying comments, basis for autoeval

AutoEval framework for data generation and model upgrading.

Master RAG system development and validation. This project uses MongoDB Atlas & Azure OpenAI, showing how to build, test, and evaluate AI applications with autoevals for measurable performance.

code for the paper "Adaptive Prediction-Powered AutoEval with Reliability and Efficiency Guarantees"

Streamlit app that uses LastMile AI's AutoEval evaluators fine-tuned on gemini and openai responses

This is a attack method for autoeval which is aim to make the test set accuracy from low to high. Our attack method is based on the confidence method which is a branch of autoeval.

AutoEval is a smart web-based academic analytics system that automates result classification, attendance detection and analysis from online assessment data. It enables teachers to upload Excel results, instantly filter branch/division-wise records, generate insights, and export structured reports reducing manual effort, errors, and analysis time.