Private local AI with a CPU-first baseline
LocalBox lets you run AI language models locally on your computer and chat with them through a simple browser interface. Your prompts and responses are processed on your machine.
Download a supported GGUF model, launch LocalBox, and start chatting. LocalBox includes a llama.cpp CPU runtime and a local OpenAI-compatible API for connecting other applications.
Features
- Local AI chat in your browser
- CPU-first inference with no dedicated GPU required
- Built-in model downloading and checksum verification
- Hardware-aware configuration recommendations
- Local model storage
- An OpenAI-compatible API
Requirements
- An x86_64 (amd64) computer
- At least 4 GiB RAM for the Lite model; 8 GiB recommended
- Additional memory and disk space for larger models
- Internet access to download models
Models are downloaded separately. Performance depends on your processor, available memory, model, and context size. Keep LocalBox running while using its browser interface.
Published by Menayer. Created by Amr El Menayer.
Support: Support@menayer.com
Website: https://www.menayer.com