ai-shifu/ChatALL

Concurrently chat with ChatGPT, Bing Chat, bard, Alpaca, Vincuna, Claude, ChatGLM, MOSS, iFlytek Spark, ERNIE and more, discover the best answers

Summary

ChatALL is a cross-platform Electron desktop client that sends a single prompt to multiple AI chat services at the same time and displays each response side by side. Built with JavaScript, Vue 3, and Vuetify 3, it targets Windows, macOS, and Linux under the Apache-2.0 license.

What it is useful for

ChatALL is useful when you need to compare answers from several large language models without opening a dozen browser tabs. It works against more than 30 bots, including ChatGPT, Claude, Gemini, Groq Cloud, xAI Grok, Cohere Command R, Llama 2, Vicuna, ChatGLM2, and the iFLYTEK and Baidu models, mixing API-backed and web-scraped connections in one queue.

Scenarios called out in the README include:

• Researchers comparing strengths and weaknesses of different LLMs across fields.
• LLM application developers debugging prompts and picking a foundation model.
• Power users running the documented "quick-prompt" mode, which fires the next prompt before the previous request finishes, so you can iterate on phrasing faster.

A documented limitation is that web-connected bots (no official API) rely on reverse-engineered interfaces and frequently break when providers change their sites. The README recommends API-backed bots for dependable work. Other features include a switchable one-, two-, or three-column view, local-only chat history, dark mode, proxy settings, multi-chat sessions, and a prompt manager.

How engineers can use it

For end users, the README points to the GitHub releases page for installers. On macOS, brew install --cask chatall is documented. On Windows, run the *-win.exe installer; on Debian-based Linux, use the .deb; on Arch, install the AUR chatall-bin package; elsewhere, run the .AppImage.

For development, clone the repo and run:

npm install
npm run electron:serve

To package a build, the README documents npm run electron:build for the current platform and npm run electron:build -- -wml --x64 --arm64 for all targets. Contributing a new bot is supported via a linked wiki guide.

Caveats to keep in mind: you must supply your own accounts or API tokens, and the project collects anonymous usage statistics such as prompt length, response length, and which answers you highlight (the content itself is not collected). Chat history and login data are stored locally in OS-specific folders, for example ~/.config/chatall/ on Linux. If something breaks, the README's troubleshooting steps include Ctrl/Cmd+R refresh, relogging through the settings button, and as a last resort deleting the local config directory to reset the app.