Overview
Airweave connects to your apps, tools and databases, continuously syncs their data, and exposes the result through a single LLM-friendly search interface. Instead of each agent reaching into Slack, Notion, Postgres and Google Drive on its own, agents make one request to Airweave and get back relevant, grounded context drawn from every connected source.
It positions itself as shared retrieval infrastructure that sits between data sources and AI systems, taking on authentication, ingestion, syncing, indexing and retrieval. The argument is one of duplication: those four concerns get rebuilt, badly, once per agent or integration, and they are the parts most likely to rot when a source's API or auth changes.
The flow is: connect your apps, databases and documents through one of 50+ integrations; Airweave syncs, indexes and exposes them through a unified retrieval layer; agents query it via SDKs, the REST API, MCP, or native integrations with popular agent frameworks; and they retrieve grounded context on demand. It runs self-hosted via Docker Compose, or as a hosted service at app.airweave.ai.
What it does
- 50+ integrations across apps, databases and document stores, kept continuously in sync
- Handles authentication, ingestion, syncing, indexing and retrieval as one layer, rather than per-agent pipelines
- A single unified, LLM-friendly search interface returning grounded context from multiple sources per request
- Reachable from agents via Python/TypeScript SDKs, a REST API, MCP, or native integrations with popular agent frameworks
- Self-hostable with Docker Compose through a single start script, or usable as hosted Airweave Cloud
Getting started
Self-hosting is a clone and one script; the script writes .env, generates the required secrets and brings up every service with health checks. Docker and docker-compose are required.
Clone and start
start.sh creates .env from .env.example, generates ENCRYPTION_KEY and STATE_SECRET, starts all services with health checks, and optionally prompts for OpenAI/Mistral API keys.
git clone https://github.com/airweave-ai/airweave.git
cd airweave
./start.shOpen the app
The first run can take two to three minutes while every service becomes healthy. The UI is then served on port 8080.
# http://localhost:8080Manage the stack
The same script restarts services, brings up the backend alone, or tears everything down.
./start.sh --restart
./start.sh --skip-frontend
./start.sh --destroyConnect sources and query
Add your apps, databases and documents from the UI, then point agents at Airweave through the SDKs, REST API or MCP. The API reference and integration guides are at docs.airweave.ai. A hosted option is available at app.airweave.ai if you would rather not run it yourself.
Commands and code are distilled from the project's own documentation — always check the official repo for the latest.
When to use it
- Give an agent one search endpoint that spans Slack, Notion, Google Drive, a CRM and a Postgres database instead of wiring each one separately
- Stop rebuilding auth-and-sync plumbing for every new agent by making retrieval a shared service
- Keep answers grounded in data that stays current, because Airweave syncs sources continuously rather than indexing once
- Expose company knowledge to any MCP-speaking client or agent framework without writing a per-framework connector
How Airweave compares
Airweave alongside other open-source rag frameworks & platforms tools AI/TLDR tracks, ranked by GitHub stars.
| Tool | Stars | What it does |
|---|---|---|
| Dify | ★ 156k | An open-source platform with a visual workflow builder for creating LLM and RAG applications without writing much code. |
| graphify | ★ 119k | Turns a folder of code, docs, PDFs and images into a local knowledge graph with tree-sitter AST parsing and Leiden communities — queryable by agents over MCP, no vector store. |
| RAGFlow | ★ 90.9k | A RAG engine built around deep document understanding that turns complex files into a grounded, citation-backed question-answering layer. |
| Pathway | ★ 62.3k | A Python framework with a Rust streaming engine that keeps ETL, real-time analytics and RAG pipelines continuously up to date as source data changes. |
| Context7 | ★ 62.2k | Context7 pulls current, version-specific documentation and code examples for any library and feeds them into your LLM, available as a CLI skill or an MCP server. |
| LightRAG | ★ 39.7k | A graph-based RAG system that builds an entity-and-relationship knowledge graph for fast retrieval and easy incremental updates. |
| Quivr | ★ 39.5k | Quivr is an open-source RAG framework that ingests your documents and answers questions about them, working with any LLM and any file type. |
| Airweave | ★ 6.6k | Sync your apps and databases into one search endpoint agents can query |