Find tools › Open-source projects › Spend less on AI

BitNet

Open-source project · by microsoft

Run ultra-compact 1-bit AI models on standard computer processors using very little power.

For developersPopular

What it costs

Free and open source; you may pay for the AI model it uses.

Use it to

  • Run 1.58-bit BitNet language models on standard CPU hardware
  • Generate text embeddings faster using low-bit embedding weights
  • Deploy low-power AI inference on edge devices

How to start

  1. Clone the repository recursively using git clone --recursive https://github.com/microsoft/BitNet.git.

  2. Build the project using CMake and the target compiler for your CPU architecture (x86 or ARM).

  3. Download an official BitNet model from Hugging Face and run the inference executable.

Good to know. Only works with architectures trained specifically as 1-bit/ternary BitNet models; standard FP16 models cannot simply run here without custom retraining.

Before you install

  • Found automaticallyNot reviewed by a person yet. Check it before installing.
  • License: MITYou are allowed to use it under these terms.
  • Last updated 2026-07-27

More detail

Official C++ inference framework designed to run ternary 1.58-bit language and embedding models quickly on standard CPUs and GPUs.

Standard LLMs require high-end GPUs and large memory pools; bitnet.cpp slashes memory bandwidth and energy use so you can run models on everyday computer processors.

Similar tools

Descriptions are written in plain English from each project's own documentation and refreshed by our agent; this one was last checked 2026-10-12. Spotted a mistake? · Open in Find tools · © 2026 Which AI Map