Ragas

Visit Website

Introduction

Ragas is an open-source library for evaluating LLM applications. Official docs describe it as a way to move from informal checks to systematic evaluation loops, with both built-in and custom metrics. The documentation lists metrics for retrieval-augmented generation (for example faithfulness, context precision and context recall) and for agents or tool use (for example tool-call accuracy and agent goal accuracy). It also covers test-data generation and integrations with frameworks such as LangChain and LlamaIndex.

The project is Apache-2.0 at vibrantlabsai/ragas and is installed with pip install ragas. The official documentation site is docs.ragas.io.

Pricing

The library is open source under Apache-2.0. See the official site for any hosted or enterprise offerings.