RemiAI

Self-hosted · Open source

Your local AI assistant, for your own data.

RemiAI runs entirely on your hardware: deep file system access, persistent memory, MCP tools, and a full agent system.

localhost:3000running locallyv2.4.0
RemiAI chat interface running on localhost
RemiAI dashboard, running on your machine

Everything you need, locally

RemiAI combines powerful AI capabilities with deep local integration, giving you full control over your data and workflow.

File System Integration

Read, search, and write files inside permitted directories, with granular permission controls and live file watching.

Persistent Memory

RemiAI remembers your preferences, interests, and personal details across every conversation, locally.

MCP Extensibility

Connect databases, APIs, and custom servers through the Model Context Protocol, with hundreds of additional tools.

Agent System

Spawn focused sub-agents for research, coding, and analysis, with parallel problem-solving and efficient token use.

Talk Mode

A hands-free voice interface with streaming speech-to-text, interruption support, and premium TTS.

Private by Design

Encrypted backups, no telemetry, and zero cloud dependencies. Your data never leaves your hardware.

Running in under a minute

Clone the repo, install dependencies, and start talking to your own local AI assistant.

  • Node.js 18 or newer
  • One-time signup code printed in your terminal
  • Ollama, Anthropic, OpenAI, or your own provider
quickstart

# Clone, install, and run

$git clone https://github.com/Houloude9IOfficial/RemiAI.git

$cd RemiAI

$npm install

$npm run dev

localhost:3000

Frequently asked questions

Short answers about how RemiAI works, where your data lives, and how to run it.

What is RemiAI?

RemiAI is a self-hosted AI assistant that runs entirely on your own hardware. It combines deep file system integration, persistent memory, MCP tool support, and a multi-agent system, so you can work with AI without sending your files to the cloud.

How is RemiAI different from ChatGPT or Claude?

RemiAI runs locally instead of in the cloud. Your conversations, files, and memory stay on your machine, it supports local models through Ollama, and it can read and write files inside directories you permit.

Is my data stored locally?

Yes. RemiAI is private by design: no telemetry, no cloud dependencies, and encrypted backups. Your data never leaves your hardware unless you connect an external provider or tool yourself.

Can I host RemiAI on my own server?

Yes. RemiAI is designed to be self-hosted on your own hardware. You can run it on your personal computer or a server you control, ensuring that your data remains private and secure.

Which AI models can I use?

RemiAI works with local models via Ollama, and with hosted providers such as Anthropic and OpenAI. You can also bring your own provider or any API-compatible endpoint.

What are the system requirements?

Node.js 18 or newer on macOS, Windows, or Linux. Setup takes about a minute: clone the repository, install dependencies, and run it with npm run dev.

Is RemiAI free and open source?

Yes. RemiAI is open source under the MIT license and free to self-host. You only pay for the AI provider you choose, if any.

Run your AI assistant on your own hardware.

Clone the repo, install dependencies, and start talking to your models in under a minute.