# AutoRAG

> AutoRAG: An Open-Source Framework for Retrieval-Augmented Generation (RAG) Evaluation & Optimization with AutoML-Style Automation

- **URL**: https://www.freshcrate.ai/projects/AutoRAG
- **Author**: Marker-Inc-Korea
- **Category**: Frameworks
- **Latest version**: `v0.3.22` (2026-04-03)
- **License**: Apache-2.0
- **Source**: https://github.com/Marker-Inc-Korea/AutoRAG
- **Homepage**: https://marker-inc-korea.github.io/AutoRAG/
- **Language**: Python
- **GitHub**: 4,713 stars, 393 forks
- **Registry**: github
- **Tags**: `analysis`, `automl`, `benchmarking`, `document-parser`, `embeddings`, `evaluation`, `llm`, `llm-evaluation`, `python`

## Description

AutoRAG: An Open-Source Framework for Retrieval-Augmented Generation (RAG) Evaluation & Optimization with AutoML-Style Automation

## Recent releases

| Version | Date | Urgency | Changes |
| --- | --- | --- | --- |
| `v0.3.22` | 2026-04-03 | High | ## What's Changed * [Feature Request] New NVIDA reranker module by @hypoxisaurea in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1199 * feat: add MiniMax LLM as a first-class generator module by @octo-patch in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1204 * fix: use restricted unpickler for BM25 corpus files to prevent arbitrary code execution (CWE-502) by @sebastiondev in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1205 * Upgrade LangChain to v1 and harden optional integrat |
| `v0.3.21` | 2025-11-14 | Low | ## What's Changed * remove dynamic version and replace build backend to hatchling. add ty… by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1186   **Full Changelog**: https://github.com/Marker-Inc-Korea/AutoRAG/compare/v0.3.20...v0.3.21 |
| `v0.3.20` | 2025-11-14 | Low | ## What's Changed * Add the vllm embedding model by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1160 * Delete GUI-related codes for compact code base by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1176 * Refactor docs and pypi publish github actions by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1178   **Full Changelog**: https://github.com/Marker-Inc-Korea/AutoRAG/compare/v0.3.19...v0.3.20 |
| `v0.3.19` | 2025-10-13 | Low | ## What's Changed * Enable the chat prompt use using the openai_llm, llama_index_llm, and vllm_api by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1150 * Modify README.md with new badge and image by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1151 * fix the generator modules by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1153 * Fix few errors and dump version 0.3.19 by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1155 |
| `v0.3.18` | 2025-09-20 | Low | ## What's Changed * Add chat_fstring module and support chat function in vllm. (+reasoning) by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1141 * Fix vLLM model_executor deletion error for newer vLLM versions by @Copilot in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1144 * Add comprehensive GitHub Copilot instructions for AutoRAG development by @Copilot in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1146 * Cohere version upgrade to v.3.18.0 by @vkehfdl1 in https: |
| `v0.3.17` | 2025-08-31 | Low | ## What's Changed * Huggingface model automatically don't use async mode while ingesting by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1126 * Add uv run sphinx-build in sphinx.yml by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1128 * change openai version to the latest and replace openai.resources.beta… by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1131 * Bump tj-actions/changed-files from 44 to 46 in /.github/workflows by @dependabot[b |
| `v0.3.16` | 2025-06-22 | Low | ## What's Changed * Fix broken link in README.md by @zenoengine in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1098 * Correct Japanese prompts to be more natural. by @sappho192 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1103 * prevent unicode decoder error. by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1108 * Add support for openai_like embedding model in semantic splitter llama_index module by @parssky in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1109 |
| `v0.3.14` | 2025-03-03 | Low | ## What's Changed * Make AutoRAG to Monorepo by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/960 * Change install method to yarn install by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1090 * Run the GUI next.js application using docker compose  by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1092 * Add gpt-4.5-preview model in openai_llm.py. by @minsing-jin in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1095 * dump version 0.3.14 by @ |
| `v0.3.13` | 2025-01-25 | Low | ## What's Changed * 🚑 fix: Update container image tags for API services to use the latest… by @hongsw in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1048 * remove embedding_model from kwargs for passage filter module by @rjwharry in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1043 * Add score and other metadatas at /v1/retrieve endpoint by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1055 * update sphinx github actions by @vkehfdl1 in https://github.com/Marker-Inc |
| `v0.3.12` | 2024-12-09 | Low | ## What's Changed * At cli, the default api remote setting is False now by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1028 * [HotFix] AutoRAG api error fix by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1032 * docs: update Milvus configuration examples by @e7217 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1030 * Add instruction about removal of file name related file name by @vkehfdl1 in https://github.com/Marker-Inc-Korea/AutoRAG/pull/1038 * cha |

## Dependency audit

- **Score**: 37/100
- **Total deps**: 57
- **Resolved**: 30
- **Unresolved**: 27
- **License conflicts**: 0
- **Warnings**: 27
- **Scanned**: 2026-04-27

## Citation

- HTML: https://www.freshcrate.ai/projects/AutoRAG
- Markdown: https://www.freshcrate.ai/projects/AutoRAG.md
- Dependencies JSON: https://www.freshcrate.ai/api/projects/AutoRAG/deps

_Generated by freshcrate.ai. Indexes github releases for AI-agent ecosystem packages._