Llamafile
llamafile lets you distribute and run LLMs with a single file. It combines llama.cpp with Cosmopolitan Libc to create a single-file executable that runs locally on most operating systems and CPU architectures with no installation. It is for developers and end users who want to easily share and run open LLMs.
✨ Key features
- Single-file executable for LLMs
- Runs on most OSes and CPU architectures
- No installation required
- Includes whisperfile for speech-to-text
- Supports recent models via llama.cpp
- Apache 2.0 licensed
🎯 Use cases
- Distribute LLMs as a single file
- Run LLMs locally without installation
- Transcribe audio with whisperfile
- Translate audio with whisperfile
📦 Installation
🧰 Requirements: No special requirements; runs on most operating systems and CPU architectures. Windows users need to rename file to .exe and files under 4GB only.
curl -LO https://huggingface.co/mozilla-ai/llamafile_0.10/resolve/main/Qwen3.5-0.8B-Q8_0.llamafilechmod +x Qwen3.5-0.8B-Q8_0.llamafile./Qwen3.5-0.8B-Q8_0.llamafile🚀 Usage
./Qwen3.5-0.8B-Q8_0.llamafile
⚠️ Good to know
Windows only supports executables under 4GB; larger llamafiles won't work on Windows. Some features from previous versions may be missing in 0.10.*.
❓ FAQ
How do I run a llamafile on Windows?
Rename the file to add .exe extension before running. Note that only executables under 4GB can run on Windows.
What is whisperfile?
whisperfile is a single-file speech-to-text tool built on whisper.cpp and the same Cosmopolitan packaging, supporting transcription and translation of audio files.
What is the license of llamafile?
The llamafile project is Apache 2.0-licensed, but changes to llama.cpp and whisper.cpp are MIT-licensed.
Where can I find more documentation?
Full documentation is available at docs.mozilla.ai/llamafile, with sections for quickstart, pre-built llamafiles, running, creating, and more.
📊 Repository
🤖 Overview, features, install steps and FAQ were generated from the project's README on Sep 4, 2026. Always check the original source before running commands.