Open Source Alternatives
  • Alternatives
  • Categories
    Categories by size1,283 projects
    • Developer Tools660
    • AI & Machine Learning543
    • Business & Productivity307
    • Web Development219
    • Security & Monitoring198
    • Communication & Collaboration168
    • IT Management164
    • Design & Creative142
    • LLMOps & AI Tooling140
    • DevOps & CI/CD137
    Browse all 21 categoriesCloud & Hosting, Product & Project Management and 9 more
  • Blog
  • Advertise
SearchNotion, Slack, password manager.../
Submit
Open Source Alternatives
  • Alternatives
    • Developer Tools660
    • AI & Machine Learning543
    • Business & Productivity307
    • Web Development219
    • Security & Monitoring198
    • Communication & Collaboration168
    • Browse all 21 categories
  • Blog
  • Advertise
Submit your project

New open source projects by email

The projects we add and every new guide. Free, unsubscribe any time.

Open Source Alternatives LogoOpen Source Alternatives

Handpicked Open Source Alternatives to Paid Softwares

Add Open Source Alternatives as a preferred source on Google

Product

  • Categories
  • Tag
  • Advertise

Resources

  • Blog
  • Collection
  • Submit

Company

  • Privacy Policy
  • Terms of Service
  • Refund Policy
  • Sitemap

Alternatives

  • Claude Code
  • Jira
  • Notion
  • Slack
  • Linear
  • Wispr Flow
  • All alternatives
© 2026 Open Source Alternatives. All rights reserved.
Open SourceAlternatives
Home/Categories/AI & Machine Learning/AutoRAG
icon of AutoRAG

AutoRAG

Open source alternative to Ragie AI

  • 5.9K stars
  • Apache-2.0
  • Active this week
  • Last commit today

Last verified Nov 17, 2026

Search your files, messengers, and cloud drives with a local-first AI librarian agent that cites every answer to its source passage.

Visit websiteGitHub5.9K stars
image of AutoRAG
TL;DR

AutoRAG Agent is a local-first personal librarian that searches your files, messengers, and cloud drives with hybrid keyword and semantic search, citing every answer to its source passage. It replaces managed RAG services like Ragie AI for personal and team knowledge retrieval, with the entire semantic indexing engine running on your machine under Apache-2.0. Best for knowledge workers and developers who want cited search across Slack, Discord, Google Drive, Obsidian, and local files without sending data to a cloud provider.

Read the full breakdownHide the full breakdownWho it is for, how it works, trade-offs, the quick start and the FAQ
who it's for

Who AutoRAG is for#

Knowledge workers searching files, messages, and cloud drives

AutoRAG indexes Slack, Discord, Mail, Google Drive, Dropbox, Obsidian, and local files in one search layer, returning cited passages instead of unverifiable summaries. Useful when you need to find a decision made in a team chat three months ago alongside the document that led to it.

Skip if:

Your work is entirely within one application (Notion-only or Google Workspace-only). A tool-native search is simpler and faster for a single-source workflow.

Developers wiring file search into Claude Code or Codex

The AutoRAG Lite MCP server plugs into any MCP-capable coding agent and gives it search over your indexed files during a session. Install by pasting the autorag-lite-setup skill prompt into your agent; it asks before touching anything and verifies with a real search after setup.

Skip if:

You need search over a large shared document corpus for a production application, not personal files. A managed RAG API like Ragie AI is a better fit when you are building a customer-facing app that needs serverless ingestion and scaling.

Teams self-hosting search over confidential documents

The local embedding model means no file contents leave the machine during indexing. Encrypted key storage and no mandatory cloud component make it suitable for teams with confidentiality requirements that rule out sending documents to a managed cloud indexer.

Skip if:

Your organization requires centralized access controls, audit logging, or multi-user permissions enforced by a server. AutoRAG is a personal and small-team tool; enterprise document management needs a different architecture.

the problem

The problem it solves#

Your files live in dozens of apps: a document on Google Drive, a decision in Slack, a code snippet in GitHub, a note in Obsidian. Searching across them means opening each one separately, remembering keywords, and getting no citation for where an answer came from. Cloud-based search tools index your content on their infrastructure, raising privacy concerns and creating ongoing costs.

The retrieval quality problem compounds this. Standard keyword search misses documents that use different terms for the same concept. Vector-only search loses exact matches. Most tools return a summary with no trail back to the original passage, so you cannot verify the answer or find what you need to read next. Teams that try to build their own retrieval layer with RAG pipelines spend weeks on infrastructure instead of on their actual work.

how AutoRAG solves it

How it solves it#

Hybrid Search with Reranker

Combines keyword matching with semantic (meaning-based) search and a reranker to surface documents that match what you mean even when your words differ from the content. Every answer links back to the specific passage in the source file, so retrieval is verifiable rather than opaque.

Multi-Source Indexing

Indexes local files alongside Slack, Discord, Telegram, WhatsApp, KakaoTalk, Mail, GitHub, Obsidian, Google Drive, iCloud Drive, OneDrive, Dropbox, and past coding-agent sessions (Claude Code, Codex, and 18 others). Data stays in each app's own store; AutoRAG reads it without copying it to a central server.

Local Semantic Embeddings

Semantic indexing runs on MinSync's local Qwen3-Embedding-0.6B engine (1024 dimensions), which AutoRAG provisions and runs on your machine. Your file contents are never sent to a remote embedder unless you explicitly configure a remote minSync.embedder.

MCP Server for Coding Agents

Ships an AutoRAG Lite MCP server so coding agents (Claude Code, Codex, Cursor) can query your indexed files during a session. Install via the autorag-lite-setup skill by pasting a single prompt into your agent; it asks before touching anything.

Duplicate Detection

Stacks proposal-final, proposal-v3, and proposal (1) copies into one row, tagged Latest, Similar, or Identical using the dupey library. Keeps the file list readable when the same document exists in multiple versions across folders.

Desktop App and Terminal UI

Ships as the Dazzi desktop app for Mac and Windows (file browsing plus AI search in one window), a terminal UI via autorag tui, a CLI, a TypeScript API, and a messenger integration for Telegram and Discord. The same retrieval engine runs in every surface.

strengths · trade-offs

Strengths and trade-offs#

Strengths

  • Apache-2.0 with no mandatory cloud componentAutoRAG's Apache-2.0 license lets you run it on your own infrastructure, modify it, and use it commercially without restriction. Unlike managed RAG services that require sending your documents to a third party, the default configuration keeps embeddings and search entirely on your machine.
  • Local embedding with no remote calls by defaultSemantic indexing uses MinSync's local Qwen3-Embedding-0.6B model provisioned on your machine. File contents stay offline unless you explicitly configure a remote embedder, which matters for teams handling confidential documents or operating in air-gapped environments.
  • Multi-surface delivery from one engineThe same librarian engine runs as a desktop app (Dazzi), terminal UI, CLI, Telegram and Discord messenger bot, MCP server, and TypeScript API. You can query your indexed files from whichever surface fits the moment without re-indexing.
  • Actively maintained since January 2024The repository's last push was on 2026-11-16, with 515 forks and 5,899 GitHub stars. New feature work targets AutoRAG Agent 2.0, while the legacy Python RAG AutoML tool receives bug fixes and continues on PyPI in maintenance mode. Bug reports and feature discussions are tracked openly in GitHub Issues and Discussions.

Trade-offs

  • -Node.js 24 and Bun 1.2 requiredThe TypeScript engine requires Node.js 24 or later and Bun 1.2 or later. Teams on older Node.js environments need to upgrade their runtime before using AutoRAG, which may conflict with other project dependencies that have not yet moved to Node.js 24.
  • -Desktop app (Dazzi) is Fair Source, not Apache-2.0The Dazzi desktop app (NomaDamas/AutoRAG-app) is licensed under FSL-1.1-MIT (Fair Source), not Apache-2.0. FSL-1.1 permits running it yourself, internal team use, and modification, but restricts offering it as a competing managed service. The core AutoRAG Agent library in this repository remains Apache-2.0.
  • -Name clash with legacy AutoRAG AutoML toolAutoRAG 2.0 is a personal librarian agent, while the original AutoRAG was a Python-based RAG pipeline AutoML tool. The legacy tool remains in the repository's legacy/ directory and continues on PyPI via pip install AutoRAG in maintenance mode. Teams searching for the AutoML pipeline optimizer may land on this repository expecting different functionality.
versus alternatives

AutoRAG vs alternatives#

AutoRAG vs Ragie AI

AutoRAG Agent and Ragie AI both address the problem of connecting documents to language models, but they serve different audiences and architectures. Ragie AI is a fully managed RAG-as-a-Service for developers building AI applications; AutoRAG is a local-first personal librarian for individuals and small teams searching their own files and apps.

FeatureAutoRAGRagie AI
LicenseApache-2.0Proprietary
Self-hostingYes, local by defaultNo (managed cloud only)
Primary use casePersonal/team file searchDeveloper API for AI apps
EmbeddingLocal Qwen3-Embedding-0.6BManaged cloud embedder
Supported sourcesFiles, 10+ messengers, cloud drivesDocuments via REST API
LLM integrationBring your own key or Dazzi ModelsCloud-managed
PricingFree (Apache-2.0, self-hosted)Commercial SaaS

AutoRAG is the better choice when you need local-first search over personal files, messengers, and cloud drives without sending your data to a managed service. Its Apache-2.0 license and local embedding engine suit privacy-conscious individuals, developers who want to give coding agents access to their own files via MCP, and small teams whose search needs are personal rather than application-layer.

Ragie AI is worth considering when you are building a customer-facing AI application that requires production-grade serverless ingestion, multimodal support, and clean REST APIs. Its fully managed service eliminates infrastructure management, which matters when indexing is a background service for an application rather than a personal search tool. Ragie AI charges as a commercial platform; AutoRAG runs for free on your own hardware.

install · quick start

Quick start#

bash
Install AutoRAG using the npm package on Node.js 24 or later.

```bash
npm install @autorag/librarian
```
tech stack · detected from GitHub

What it's built on#

Languages
JavaScriptPythonTypeScript
Frameworks
FastAPI
frequently asked

FAQ#

Is AutoRAG Agent free to use?

The core AutoRAG Agent library and CLI are Apache-2.0 licensed and free to run on your own machine. The Dazzi desktop app is licensed under FSL-1.1-MIT (Fair Source), which permits personal and team use but restricts offering it as a competing managed service. A hosted Dazzi Models plan is available for the LLM inference layer if you do not want to supply your own API key.

Does AutoRAG send my files to the cloud?

By default, no. Semantic indexing runs on a local Qwen3-Embedding-0.6B model that AutoRAG provisions on your machine. File contents are not sent to a remote embedder unless you explicitly configure a remote minSync.embedder. Sign-in tokens and API keys are stored in the system's encrypted store, not in a config file.

What is the difference between AutoRAG Agent and the legacy AutoRAG?

The repository now hosts AutoRAG Agent 2.0, a personal librarian for files, messengers, and cloud drives. The original AutoRAG was a Python-based RAG AutoML tool for finding optimal RAG pipelines for datasets. The legacy tool is in maintenance mode in the legacy/ directory and continues to receive bug fixes and PyPI releases via pip install AutoRAG. New feature development is focused entirely on AutoRAG Agent 2.0.

Which datasources does AutoRAG support?

AutoRAG indexes local files, Slack, Discord, Telegram, WhatsApp, KakaoTalk, Mail, GitHub (and GitHub Gists), Obsidian, Google Drive, iCloud Drive, OneDrive, Dropbox, browser history, RSS, and past coding-agent sessions from Claude Code, Codex, and others via the wombat integration. Photos are also supported. Cloud drives are searched through their synced folders. The full list is in docs/datasource-skills.md in the repository.

How do I install AutoRAG Agent?

The quickest path is the agent-assisted setup: paste the autorag-setup skill prompt (in the repository's skills/ folder) into Claude Code, Codex, or Cursor and the agent handles installation, folder selection, and health check. For manual setup, the npm package is @autorag/librarian (Node.js 24+ required). The Dazzi desktop app is available for Mac and Windows at dazziapp.com without a command-line setup.

also worth a look

Similar open-source tools#

RAGMeUp

Alternative toVectara

Open source RAG framework for any dataset or LLM

  • 680 GitHub stars
  • Apache-2.0Licence Apache-2.0

RAG-Anything

Alternative toRagie AI

Comprehensive multimodal document processing framework

  • 24,517 GitHub stars
  • MITLicence MIT

Atomic Agent

Alternative toGrok Bot

Local-first AI agent that runs on your machine, drives real tools.

  • 3,680 GitHub stars
  • MITLicence MIT

rakazo

Featured

Alternative toGrok Bot

AI teammates you own: your keys, your model, your machine.

  • 8,291 GitHub stars
  • Apache-2.0Licence Apache-2.0

Hermes Agent

Featured

Alternative toGrok Bot

Self-hosted AI agent: persistent memory, multi-channel chat, model choice.

  • 263,254 GitHub stars
  • MITLicence MIT

Ninfer

InAI & Machine Learning

Native C++/CUDA inference for Qwen3 and GLM models, self-hosted.

  • 3,356 GitHub stars
  • Apache-2.0Licence Apache-2.0
by category

Open source tools in these categories#

543

AI & Machine Learning

freeCodeCamp, REA (Reverse Engineer Anything), OpenClaw
140

LLMOps & AI Tooling

Hermes Agent, OpenCode, MarkItDown
660

Developer Tools

REA (Reverse Engineer Anything), Skills for Real Engineers, superpowers
307

Business & Productivity

OpenClaw, Excalidraw, RustDesk

Sponsors

Inbox Zero

Alternative toSuperhuman

AI email assistant that drafts replies and clears your inbox

  • 13,025 GitHub stars
  • AGPL-3.0Licence AGPL-3.0
  • #Email

rakazo

Alternative toGrok Bot

AI teammates you own: your keys, your model, your machine.

  • 8,291 GitHub stars
  • Apache-2.0Licence Apache-2.0
  • #AI Agents

SnitchFeed

Live monitoring on LinkedIn, X, Reddit & more. Agentic social listening for startups.

Marketing & Growth

Advertise your product

New open source projects by email

The projects we add and every new guide. Free, unsubscribe any time.

Find open source alternatives to paid software

/
  • Claude Code to OpenCode
  • Jira to Windshift
  • Notion to AppFlowy
  • Slack to Mattermost
  • Linear to GitLab
  • Wispr Flow to Handy
  • Arc to Zen Browser
  • CapCut to Concat
  • Grok Bot to rakazo
  • Superhuman to Inbox Zero
  • Discord to Stoat
  • Airtable to AITable
  • Datadog to SigNoz
  • Muse to OpenClaw
  • Railway to Coolify
  • GitHub Copilot to Cline
  • Confluence to Wiki.js
  • Microsoft Teams to Rocket.Chat
  • 1Password to KeePassXC
  • Firebase to Supabase
  • Pinecone to Qdrant
  • Zapier to n8n
  • Devin AI to OpenHands
  • Asana to Leantime
  • Claude Code to OpenCode
  • Jira to Windshift
  • Notion to AppFlowy
  • Slack to Mattermost
  • Linear to GitLab
  • Wispr Flow to Handy
  • Arc to Zen Browser
  • CapCut to Concat
  • Grok Bot to rakazo
  • Superhuman to Inbox Zero
  • Discord to Stoat
  • Airtable to AITable
  • Datadog to SigNoz
  • Muse to OpenClaw
  • Railway to Coolify
  • GitHub Copilot to Cline
  • Confluence to Wiki.js
  • Microsoft Teams to Rocket.Chat
  • 1Password to KeePassXC
  • Firebase to Supabase
  • Pinecone to Qdrant
  • Zapier to n8n
  • Devin AI to OpenHands
  • Asana to Leantime