This Emacs package provides a unified shell interface for interacting with multiple Large Language Models (LLMs) including ChatGPT, Claude, Gemini, and Ollama. It targets Emacs users who want to integrate LLM capabilities directly into their text editing workflows, offering features like inline code modification, execution via Org Babel, and image description.

How It Works

The package leverages Emacs's comint shell interface, allowing users to interact with LLMs in a familiar terminal-like environment within Emacs. It supports a wide array of LLM providers by abstracting API interactions, enabling users to switch between models seamlessly using a dedicated command. A key feature is the "compose" buffer, which facilitates crafting more complex prompts by providing an editable scratchpad before submission.

Quick Start & Requirements

• Install: Via MELPA or Straight.
• Prerequisites: Emacs. API keys for cloud-based LLM services (OpenAI, Anthropic, Google, etc.) are required. Ollama requires local installation.
• Setup: Configure API keys via Emacs Lisp variables or password managers.
• Docs: https://github.com/xenodium/chatgpt-shell

Highlighted Details

• Supports multiple LLMs: ChatGPT, Claude, DeepSeek, Gemini, Kagi, Ollama, Perplexity.
• Compose buffer for detailed prompt crafting and interaction management.
• Inline modification confirmation via diffs.
• Execution of LLM-generated code snippets using Org Babel.
• Image description capabilities (currently ChatGPT only).

Maintenance & Community

The project is actively developed by xenodium, with support encouraged via GitHub Sponsors. Community interaction is available through various social media platforms linked in the README.

Licensing & Compatibility

The README does not explicitly state a license. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

Image query support is limited to ChatGPT. The project is described as "in development," suggesting potential for ongoing changes and undiscovered issues. The license is not clearly stated, which may impact commercial adoption.