Chatbox Goes Open Source: A Desktop Client for ChatGPT, Claude, and More

If you’ve been juggling multiple AI chat interfaces—ChatGPT in one tab, Claude in another, and maybe Gemini or Grok somewhere else—you know the pain. Each web app has its own quirks, UI changes, and login sessions. That’s where Chatbox comes in. It’s a single desktop client that wraps multiple AI providers into one consistent, local app. And now, it’s open source again.

The project was recently re-licensed to AGPL-3.0, meaning you can grab the code, inspect it, modify it, and even self-host if you want. It’s built with Electron (like many desktop apps), but the team kept it lightweight and focused: no bloat, no webview dependencies, just a clean chat interface that talks to APIs directly.

What It Does

Chatbox is a desktop client for AI chat services. Think of it like a unified inbox for your conversations across:

• ChatGPT (GPT-3.5, GPT-4, GPT-4o)
• Claude (Sonnet, Opus)
• Gemini
• Groq (fast inference)
• Local models via Ollama or LM Studio

You configure API keys for each provider once, then switch between models with a dropdown. The app stores all conversations locally (SQLite database), so nothing leaves your machine unless it’s going to the API.

Why It’s Cool

1. It’s open source (AGPL-3.0).
The code is on GitHub, so you can verify it’s not exfiltrating data. You can also fork it, add features, or build your own custom AI client. The re-licensing happened recently, and the community is activelly contributing.

2. Multiple providers in one window.
You don’t need to log into 4 different tabs. Want to compare responses from GPT-4 and Claude on the same prompt? Just switch the model and keep the same conversation. Each message notes which model generated it.

3. Local-only data.
All chat histories are stored in a local database. No cloud sync (though some have asked for it—it’s on the roadmap). If you care about privacy, this is a big plus.

4. Prompt templates and presets.
You can create reusable prompts (e.g., “write a commit message”, “explain this code”) and save them as templates. Useful for developers who use AI as a coding assistant.

5. Runs everywhere.
Windows, macOS, Linux. It’s an Electron app, so it’s cross-platform. Installation is a single download or a brew install chatbox on macOS.

How to Try It

• Download the latest release: github.com/chatboxai/chatbox/releases
• Mac users: brew install --cask chatbox
• Linux: Grab the AppImage or .dmg from the releases page.
• Build from source: Clone the repo, run npm install && npm start

You’ll need API keys for each service you want to use. The app will guide you through setting them up on first launch. No registration required beyond that.

Final Thoughts

Chatbox won’t replace your terminal or IDE, but it is a solid companion for anyone who regularly uses multiple AI chat models. The open source re-release is a smart move—it builds trust and invites contributions. For developers, it’s also a neat case study in building a desktop client around public APIs. If you’ve ever wanted to compare model outputs side-by-side without losing your mind, this is worth trying.

Follow us for more open source projects: @githubprojects