RLLM

A local first LLM chat UI with end to end encrypted syncing between devices.

There is a demo available, currently it has syncing/account enabled but it will be disabled later when all testing is done. Demo Link

Screenshots

Coming Soon…

Use Demo Link for now

Features

• Local first. Access your chat history offline. Use LLMs even if your web server is down but your LLM provider is up.
• End to end encrypted syncing between devices
• Use any provider that has an openai chat completions API compatible endpoint (Openrouter, llama-server, ollama, vllm, etc...)
• MCP Tools Support
• Client side RAG for PDF/EPUB files
• Image input support
• Cors Proxy Support for MCP servers and LLM providers that don't set CORS headers correctly
• Easy Install with Docker Compose

Roadmap

• Settings - system prompt, temperature etc…
• Presets - Save/Load settings, enabled MCP tools, provider and model as a named preset
• MCP Resources / Prompts
• Rewrite Sync Server in Rust or Go

Installation

docker-compose

1. clone this repo

git clone https://github.com/vanillacode314/rllm

2. (optional: if you want syncing support) uncomment sync-server and database in docker-compose.yml and copy env.example to .env

cp env.example .env

3. (optional: if you want cors proxy) uncomment go-cors-proxy in docker-compose.yml
4. run docker compose

docker-compose up -d --build

4. Open http://localhost:9870 in your browser

Motivation

I used to self host openwebui and librechat on my laptop, it bothered me that I couldn't access chat history on my mobile when my laptop was off or that I couldn't use external providers that were up even when my laptop was off.

Author

👤 Raqueebuddin Aziz (vanillacode314)

I freelance as a web app developer. Reach out if you like my work :)

• Website: https://raqueeb.com
• Github: @vanillacode314

Show your support

Give a ⭐️ if you like the project

📝 License

This project is GPL-3.0 licensed.