Clippy lets you run large language models (LLMs) locally on your computer while sticking with a 1990s user interface. Through Llama.cpp, it supports models in popular GGUF format (most publicly available models). Comes with one-click installation support for Google's Gemma3, Meta's Llama 3.2, Microsoft's Phi-4, and Qwen's Qwen3. A love letter and homage to the late great Clippy, assistant from Microsoft Office 1997. Character designed by illustrator Kevan Atteberry, who created 15+ potential characters for Microsoft's Office Assistants. Not affiliated, approved, or supported by Microsoft. Consider it software art or software satire. Also meant as reference implementation of @electron/llm hoping to help other Electron app developers make use of local language models.
Features simple, familiar, classic chat interface (send messages to models, get responses), batteries included with no complicated setup (just open app and chat), thanks to llama.cpp and node-llama-cpp automatic discovery of most efficient way to run models (Metal, CUDA, Vulkan, etc.), custom models/prompts/parameters (load your own downloaded models and play with settings), offline/local/free (everything runs on computer, only network request for update checks which can be disabled). Non-features: Countless little chat apps for local LLMs exist; many likely better. This project isn't trying to be your best chat bot but rather weird mix of nostalgia for 1990s technology paired with magical 2025 technology running on computers. Supports most GGUF models via Llama.cpp; find GGUF models from TheBloke or Unsloth on HuggingFace. Thanks to Microsoft, Kevan Atteberry for Clippy, Jordan Scales for Windows 98 design, Pooya Parsa for extracting Clippy spritesheet frames, and node-llama-cpp for squeezing llama.cpp into Node.js. NOASSERTION license.
Use Cases:
Clippy (https://felixrieseberg.github.io/clippy/) let's you run a variety of large language models (LLMs) locally on your computer while sticking with a user interface of the 1990s. Through Llama.cpp, it supports models in the popular GGUF format, which is to say most publicly available models. It comes with one-click installation support for Google's Gemma3, Meta's Llama 3.2, Microsoft's Phi-4, and Qwen's Qwen3.
It's a love letter and homage to the late, great Clippy, the assistant from Microsoft Office 1997. The character was designed by illustrator Kevan Atteberry, who created more than 15 potential characters for Microsoft's Office Assistants. This app is not affiliated, approved, or supported by Microsoft. Consider it software art. If you don't like it, consider it software satire.
It is also meant to be a reference implementation of @electron/llm (https://github.com/electron/llm), hoping to help other developers of Electron apps make use of local language models.
node-llama-cpp, the app will automatically discover the most efficient way to run your models (Metal, CUDA, Vulkan, etc).Countless little chat apps for local LLMs exist out there. Many of them are likely better - and that's okay. This project isn't trying to be your best chat bot. I'd like you to enjoy a weird mix of nostalgia for 1990s technology paired with one the most magical technologies we can run on our computers in 2025.
Clippy supports (thanks to Llama.cpp) most GGUF models. You can find GGUF models in plenty of online sources - I tend to go with models quantized by TheBloke (https://huggingface.co/thebloke) or Unsloth (https://huggingface.co/unsloth).
Thanks to:
AI assistant for tmux that reads terminal panes for context. Features Observe, Prepare and Watch modes with command execution, safety scoring and knowledge base support. Works with OpenRouter, OpenAI and Azure.
Desktop app to manage MCP servers on Windows and macOS. Organize servers into Projects and Workspaces. Connects with Claude, Cline, Windsurf and Cursor. All data stored locally for privacy.
Raycast-compatible launcher for Linux with command palette, Raycast Store extensions, smart calculator, clipboard history and snippets. Built with Tauri and Rust. Available as AppImage.
Android app for running Stable Diffusion with Snapdragon NPU acceleration. Supports text-to-image, image-to-image and inpainting with custom models, LoRA and upscaling. Works on Android 5.0 and newer.
Fast SQL client for PostgreSQL, MySQL and SQL Server with AI assistant that converts natural language to queries. Features Monaco editor, ERD diagrams, query plans and inline editing. Built with Electron and React.
Convert websites into desktop apps with Electron. Features multi-account support, global hotkey switching, custom JavaScript injection and portable packaging for Windows, macOS and Linux.