Switch between ChatGPT, Claude, and Gemini from one desktop app without opening multiple browser tabs.
Run a completely offline AI assistant on your own hardware using Ollama with zero data leaving your machine.
Keep all your conversation history private and stored locally instead of on a cloud provider's servers.
Use saved prompt templates and keyboard shortcuts to speed up repetitive AI tasks.
Requires Node.js runtime and API keys for external models (ChatGPT/Claude/Gemini) or local model setup; Electron build may need platform-specific dependencies.
Chatbox is a desktop and mobile application that gives you a polished, privacy-respecting interface for interacting with large language models like ChatGPT, Claude, Google Gemini, and locally-run models through Ollama. The problem it solves is simple: the official web interfaces for these AI services are browser-only, store your conversations on the provider's servers, and lack features that power users want, such as a prompt library, keyboard shortcuts, and offline access to conversation history. Chatbox works by connecting to AI provider APIs using your own API key, which you configure in the app's settings. This means all your conversation data is stored locally on your device rather than in a cloud account. The app renders responses in real time with full support for Markdown formatting, code syntax highlighting, and mathematical notation via LaTeX, so technical and scientific output looks clean and readable. It also supports local models through Ollama, meaning you can run an entirely offline AI assistant on your own hardware with no data leaving your machine at all. You would use Chatbox if you want a single, consistent interface to switch between multiple AI providers without juggling browser tabs, if you care about keeping your conversation history private and stored locally, or if you want advanced features like saved prompt templates, message quoting, and dark mode in a native desktop application. Teams can also share an OpenAI API key through a built-in team collaboration feature. The application is built with TypeScript and runs as a native desktop app using Electron on Windows, macOS, and Linux, as well as offering native iOS and Android apps. The community edition is open-source under the GPLv3 license.
Generated 2026-05-18 · Model: sonnet-4-6 · Verify against the repo before relying on details.