Skip to content
All packages

llama.cpp

AI / ML · C/C++

  • Metadata verified
  • Licence confirmed
  • Reviewed

C/C++ runtime for running large language models locally on CPU or GPU, using quantised GGUF weights and an included HTTP server.

Repository Documentation

Runs on ordinary hardware without a GPU budget, which makes it a practical way to see how inference and quantisation actually behave.Telebem editors

Details

Use case
AI / ML
Language
C++C/C++ ecosystem
Licence
MITPermissive: Use it in anything, including closed software, if you keep the notice.The code is MIT, but the model weights you run with it are licensed separately and are often far more restrictive.
Activity
ActiveReleasing regularly
Package
Source build (also packaged as llama.cpp on Homebrew)
Tags
llm · inference · local-models · quantisation · gguf

What we checked

  • Metadata verified

    Name, package, repository and documentation checked against each other. Last run .

  • Licence confirmed

    The declared SPDX identifier was read from the repository’s own LICENSE file. Last run .

  • Reviewed

    An editor read the source and wrote the note explaining why it is listed. Last run .

These checks describe a project’s licence, metadata and release activity. They are not a security audit, and open source does not mean safe. Review any dependency before you ship it.

Also for AI / ML