chu2bard/agentbench

Evaluation framework for AI coding agents

Python1 stars0 forks

Topics

#agents#ai#benchmarks#evaluation#python#testing

What it does

Agentbench is an evaluation framework designed for AI coding agents, enabling users to define benchmarks, run agents, and collect performance metrics. This tool is crucial for developers and researchers looking to assess and improve AI coding capabilities.

Star history

Not enough history yet — 1 day(s) recorded. The daily snapshot builds this up.

Tracking

Last trending: 2026-02-21

Creator kit

Hook

Discover how Agentbench can transform the evaluation of AI coding agents and enhance your development workflow!

Content angles

Create a tutorial on setting up and using Agentbench for AI coding agent evaluation.
Discuss the importance of benchmarking AI coding agents and how Agentbench facilitates this process.
Share a case study showcasing the performance metrics collected using Agentbench with various AI coding agents.

Who should care

Developers, researchers, and enthusiasts interested in AI and coding automation.