Open source desktop application

A desktop app for when you outgrow ChatGPT.

An AI that can see your files, access live data, remember everything, and take actions—locally, privately, with any model.

  • Any model — OpenAI, Anthropic, Ollama, or local
  • Your files indexed automatically
  • 50+ connectors + any MCP server
  • Persistent memory across sessions
  • One app — no pip, no npm, no docker
Download for macOS Download for Windows
View on GitHub • macOS, Windows, Linux
RZN

Product screenshot / demo will go here

Split-pane workspace • Document viewer • Chat • Editor

Things you can't do with ChatGPT.

A desktop app can do things a browser tab can't.

MULTI-MODEL

"Claude's answer was better. Let me continue with Claude."

Send the same question to GPT-4, Claude, and Gemini simultaneously. Compare answers side-by-side. Continue the conversation with whichever model gave the best response.

  • Switch models mid-conversation
  • Side-by-side comparison
  • Use local models (Ollama, LM Studio)

Screenshot / GIF showing multi-model comparison

Screenshot / GIF showing file search

FILE INDEXING

"What did that PDF say about distributed consensus?"

Ask about a paper you downloaded three weeks ago. RZN already indexed it. No re-uploading, no 10MB limits, no expiring uploads.

  • Automatic local file indexing
  • PDFs, docs, code, markdown
  • Your data stays on your machine
50+ CONNECTORS

"Search my Google Drive for that doc."

Your cloud storage is connected. Ask the AI to find files, summarize documents, or pull context from Drive, Dropbox, or OneDrive without leaving the app.

  • Google Drive, Dropbox, OneDrive
  • Gmail, Outlook, Slack
  • Add your own via MCP

Screenshot / GIF showing connector panel

Screenshot / GIF showing code execution

CODE INTERPRETER

"Run this analysis on my local CSV."

Python interpreter built in. Analyze data, generate charts, transform files—all without uploading sensitive data to the cloud.

  • Sandboxed Python execution
  • Charts, data analysis, file transforms
  • Results stay in context

ChatGPT and Claude are excellent. But they live in a browser tab. RZN lives on your machine—with access to your files, your data, and your tools.

Why a desktop app?

Browser-based AI tools have a fundamental problem: they can't see your files. Every session starts from zero. Upload a PDF, ask questions, close the tab, and tomorrow you're uploading the same PDF again.

A desktop app changes everything. Your filesystem becomes the context. That research paper you downloaded last month? Already indexed. Your entire Documents folder? Searchable. No uploading, no size limits, no expiring links.

And because it runs on your machine, you control what stays local and what gets sent to APIs. Run embeddings locally. Use Ollama for private conversations. Execute Python code without sending your data anywhere. Your machine, your rules.

The trade-off? You download an app instead of opening a browser tab. We think that's worth it.

The pieces that make AI actually useful.

LLM

A model you choose.

Switch mid-conversation. Create profiles for different tasks. Cloud, local, or fully airgapped.

OpenAI OpenAI
Anthropic Anthropic
Google Google
Mistral Mistral
Ollama Ollama
🖥️ LM Studio
+ Any OpenAI-compatible endpoint

Context

Your files and history as context.

Local folders indexed automatically—edits picked up without re-uploading. Cloud storage connected. Emails searchable. Conversation history persists across sessions.

Local:

Folders
PDFs
Markdown
Code

Cloud Storage:

Google Drive Google Drive
Dropbox Dropbox
OneDrive OneDrive

Productivity:

Gmail Gmail
Outlook Outlook
Slack Slack
Notion Notion

Tools

Tools that read and write.

Fetch information, send messages, update documents. 50+ built-in integrations—and add your own via MCP.

Research & Discovery:

Hacker News Hacker News
Reddit Reddit
arXiv arXiv
PubMed PubMed
YouTube YouTube
Wikipedia Wikipedia

Work & Productivity:

Gmail Gmail
Slack Slack
Google Drive Google Drive
Notion Notion
Linear Linear
GitHub GitHub

Custom:

🔌 MCP
Add your own tools via Model Context Protocol

Code

Code execution with guardrails.

Python interpreter built in. Analyze data, generate charts, transform files. Runs sandboxed—results stay in context.

# Example:
import pandas as pd
df = pd.read_csv('data.csv')
df.plot()

Screenshot of code interpreter running

All of it. One app.

Get Started

Available for macOS, Windows, and Linux

Download for macOS Download for Windows Download for Linux
View on GitHub • Star us, contribute, or build from source

Setup in 2 Minutes

1

Install

Download and run the installer. Single application—no Docker, no pip, no npm.

2

Choose Your Approach

Option A: Self-Managed
  • • Add your API keys (OpenAI, Anthropic, Google, or others)
  • • Configure local models (Ollama, LM Studio) if preferred
  • • Point to your documents folder
  • • Start chatting
Option B: Managed Service (if you don't want to manage API keys)
  • • Sign up for RZN managed service
  • • One flat monthly fee
  • • We handle API keys for OpenAI, Anthropic, Google, and 50+ integrations
  • • No key management, one bill
Option C: Enterprise Self-Hosted
3

Start Using

  • • Open a document (EPUB, PDF, web page)
  • • Ask a question
  • • Take notes
  • • Everything indexed and searchable

What Happens After Download?

First launch

  • • Quick onboarding (30 seconds)
  • • Configure your first model
  • • Index your first document
  • • See how it works

Your data stays local

  • • Documents indexed on your machine
  • • SQLite database stored locally
  • • Only API calls go to cloud (if using cloud models)
  • • Full privacy and control

Open source

  • • Code available on GitHub
  • • Community-driven development
  • • Contribute features, report issues
  • • Build your own extensions via MCP

Need Help?

Documentation Discord GitHub Issues Email: support@rzn.app

Ready to stop switching between apps?

Download for macOS Download for Windows Download for Linux
View on GitHub • Read the Docs • See Features