Vane
A self-hosted AI answering engine that searches the web and cites its sources.
- Stars
- 38k
- Forks
- 4.2k
- Licence
- MIT
- Language
- TypeScript
- Last update
- 2 months ago
- Type
- App
What it is
Vane is an AI answering engine you run on your own machine. You ask a question, it searches the web through SearxNG and uses a language model you choose to write an answer with cited sources. It can use local models through Ollama or cloud providers such as OpenAI, Claude, and Groq.
- Ask a research question and get an answer with cited web sources
- Search discussions or academic papers by choosing those sources
- Upload a PDF or text file and ask questions about it
- Limit a search to specific websites, such as technical documentation
- Browse the Discover page for articles and trending content
Great for
- People who want AI-written answers with links to sources
- Users who prefer to run their own search setup instead of a hosted service
- People who want to use local models through Ollama
- Developers who want a self-hosted starting point for AI search
Not ideal for
- People who want a zero-setup hosted service, since the README centers on running it yourself with Docker or Node.js
- Users who need features the README lists as coming soon, such as Tavily and Exa search support
- People without a model provider or local LLM, since Vane needs one to produce answers
How to get it
- 1
Run with Docker
You'll need: Docker
Run with Docker (includes SearxNG)
docker run -d -p 3000:3000 -v vane-data:/home/vane/data --name vane itzcrazykns1337/vane:latestClone the repository
git clone https://github.com/ItzCrazyKns/Vane.git - 2
Run it
- Run the Docker command to start the Vane container.
- Open http://localhost:3000 in your browser.
- Enter your API keys, models, and other settings in the setup screen.
- Ask a question in the search box.
- 3
Get the most out of it
- Connect a local model through Ollama if you want answers without sending queries to a cloud provider
- Choose Speed, Balanced, or Quality mode depending on how deep you want the search to go
- Select web, discussion, or academic sources to narrow results
- Use domain search to target specific documentation sites
- Upload documents to ask questions about your own files
Downloads
Strengths
- Can run with local models so queries stay on your hardware
- Supports many AI providers and lets you mix models
- Includes a bundled SearxNG so setup can be a single Docker command
Trade-offs
- Answer quality depends on the model and provider you connect, and the README does not compare them
- Some listed features, such as Tavily and Exa support, are described as coming soon
- Connecting local models like Ollama can require network and firewall changes on Linux
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| Vane (this) | 38k | App | intermediate | 2 months ago | People who want AI-written answers with links to sources |
| anything-llm | 67.8k | App | beginner | 6 days ago | People who want a private ChatGPT-style assistant over their own PDFs and text files |
| DocsGPT | 18.4k | App | intermediate | 4 days ago | Teams that want an internal assistant over their own documents and wikis |
| Prompt-Engineering-Guide | 79.8k | Learning resource | beginner | 8 months ago | People who are new to prompting and want structured lessons with examples |
| ai-engineering-from-scratch | 76.3k | Learning resource | intermediate | 4 days ago | Students who want to understand how LLMs, retrieval, and agents work internally |
| AI-For-Beginners | 70.5k | Learning resource | beginner | 13 days ago | Beginners who want a structured, self-paced introduction to AI |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.