llama.cpp

LLM inference in C/C++

Fork notice: This is a fork of TheTom/llama-cpp-turboquant with CUDA-optimized TurboQuant KV cache quantization. See below for details.

TurboQuant CUDA KV Cache Quantization

This fork adds full CUDA support for TurboQuant (Google Research, ICLR 2026) KV cache quantization, including custom Flash Attention kernels and a tensor-core prefill path. The result: 3.5-5x KV cache compression with quality that beats q8_0.

What's implemented

turbo3 (3.25-bit) and turbo4 (4.25-bit) CUDA quantization and dequantization
FWHT rotation (Fast Walsh-Hadamard Transform) applied during quantization for near-uniform distribution
Norm correction (corrected_norm = group_norm / recon_norm) — the key to beating q8_0 quality
QJL sign correction (turbo4) — 1-bit Johnson-Lindenstrauss residual encoding
Custom Flash Attention vec kernels for both turbo3 and turbo4 K dot-product and V dequant
Prefill dequant+MMA path — bulk-dequants turbo KV to fp16 temp buffers for tensor core acceleration (turbo3 only; turbo4's QJL correction loses precision in fp16 round-trip)
Layer-adaptive modes — keep critical layers (first/last) at q8_0 for even better quality
fp16 centroid LUT in Flash Attention for decode speed
Float norm broadcast optimization for vec dequant

Benchmarks (Qwen3.5-27B Q6_K, RTX 3090 24GB)

Quality (Perplexity, lower is better)

Config	PPL	vs q8_0	KV Compression
q8_0 baseline	5.8375	—	1.0x
LA-1 turbo3	5.7690	-1.17%	3.5x
LA-5 turbo3	5.8091	-0.49%	4.2x
turbo3 uniform	5.8323	-0.09%	4.9x
turbo4 uniform	5.8186	-0.32%	3.8x

Speed

Config	Prefill (pp4096)	Decode (tg64)	vs q8_0
q8_0 baseline	1133 tok/s	31.04 tok/s	—
LA-1 turbo3	1128 tok/s (99.6%)	30.25 tok/s (97.5%)	~98%
turbo3 uniform	1125 tok/s (99.3%)	30.04 tok/s (96.8%)	~98%

Extreme Context (128K on 24GB — impossible with q8_0)

Context	Prefill tok/s	Decode tok/s	Config
65K	847	29.79	turbo3 uniform
128K	671	29.89	turbo3 uniform
128K	674	30.01	LA-5 turbo3

q8_0 OOMs beyond ~65K context on 24GB. TurboQuant enables full 128K context with constant decode speed.

Usage

Build with CUDA Flash Attention support:

cmake -B build -DGGML_CUDA=ON -DGGML_NATIVE=ON -DGGML_CUDA_FA=ON -DGGML_CUDA_FA_ALL_QUANTS=ON -DCMAKE_BUILD_TYPE=Release
cmake --build build -j$(nproc)

Run with turbo3 KV cache:

# Recommended: Layer-adaptive mode 1 (best quality, up to ~65K context)
TURBO_LAYER_ADAPTIVE=1 ./build/bin/llama-cli -m model.gguf -ctk turbo3 -ctv turbo3 -fa on

# For 128K context on 24GB GPU
TURBO_LAYER_ADAPTIVE=5 ./build/bin/llama-cli -m model.gguf -ctk turbo3 -ctv turbo3 -fa on

# Maximum compression (4.9x)
./build/bin/llama-cli -m model.gguf -ctk turbo3 -ctv turbo3 -fa on

Context length recommendations

Context Range	Config	PPL vs q8_0	Compression
Up to 65K	`TURBO_LAYER_ADAPTIVE=1` turbo3	-1.17% (better)	3.5x
65K-128K	`TURBO_LAYER_ADAPTIVE=5` turbo3	-0.49% (better)	4.2x
128K+	turbo3 uniform	-0.09% (better)	4.9x

Key findings

FWHT rotation is essential: +6.8% PPL without it
Norm correction makes turbo3 beat q8_0 — without it, quality is ~12% worse
QJL adds +0.3 PPL to turbo4 — worth keeping despite preventing MMA prefill
Decode speed gap (~3% vs q8_0) is structural: memory-bound, q8_0 benefits from native int8 dp4a on Ampere
Layer-adaptive mode 1 (first4+last4 at q8_0) gives the best quality at modest compression cost
Decode speed is constant across context lengths (30 tok/s at 4K, 65K, and 128K)

Recent API changes

Hot topics

guide : using the new WebUI of llama.cpp
guide : running gpt-oss with llama.cpp
[FEEDBACK] Better packaging for llama.cpp to support downstream consumers 🤗
Support for the gpt-oss model with native MXFP4 format has been added | PR | Collaboration with NVIDIA | Comment
Multimodal support arrived in llama-server: #12898 | documentation
VS Code extension for FIM completions: https://github.com/ggml-org/llama.vscode
Vim/Neovim plugin for FIM completions: https://github.com/ggml-org/llama.vim
Hugging Face Inference Endpoints now support GGUF out of the box! ggml-org#9669
Hugging Face GGUF editor: discussion | tool

Quick start

Getting started with llama.cpp is straightforward. Here are several ways to install it on your machine:

Install llama.cpp using brew, nix or winget
Run with Docker - see our Docker documentation
Download pre-built binaries from the releases page
Build from source by cloning this repository - check out our build guide

Once installed, you'll need a model to work with. Head to the Obtaining and quantizing models section to learn more.

Example command:

# Use a local model file
llama-cli -m my_model.gguf

# Or download and run a model directly from Hugging Face
llama-cli -hf ggml-org/gemma-3-1b-it-GGUF

# Launch OpenAI-compatible API server
llama-server -hf ggml-org/gemma-3-1b-it-GGUF

Description

The main goal of llama.cpp is to enable LLM inference with minimal setup and state-of-the-art performance on a wide range of hardware - locally and in the cloud.

Plain C/C++ implementation without any dependencies
Apple silicon is a first-class citizen - optimized via ARM NEON, Accelerate and Metal frameworks
AVX, AVX2, AVX512 and AMX support for x86 architectures
RVV, ZVFH, ZFH, ZICBOP and ZIHINTPAUSE support for RISC-V architectures
1.5-bit, 2-bit, 3-bit, 4-bit, 5-bit, 6-bit, and 8-bit integer quantization for faster inference and reduced memory use
Custom CUDA kernels for running LLMs on NVIDIA GPUs (support for AMD GPUs via HIP and Moore Threads GPUs via MUSA)
Vulkan and SYCL backend support
CPU+GPU hybrid inference to partially accelerate models larger than the total VRAM capacity

The llama.cpp project is the main playground for developing new features for the ggml library.

Models

Typically finetunes of the base models below are supported as well.

Instructions for adding support for new models: HOWTO-add-model.md

Text-only

Multimodal

Bindings

Python: ddh0/easy-llama
Python: abetlen/llama-cpp-python
Go: go-skynet/go-llama.cpp
Node.js: withcatai/node-llama-cpp
JS/TS (llama.cpp server client): lgrammel/modelfusion
JS/TS (Programmable Prompt Engine CLI): offline-ai/cli
JavaScript/Wasm (works in browser): tangledgroup/llama-cpp-wasm
Typescript/Wasm (nicer API, available on npm): ngxson/wllama
Ruby: yoshoku/llama_cpp.rb
Rust (more features): edgenai/llama_cpp-rs
Rust (nicer API): mdrokz/rust-llama.cpp
Rust (more direct bindings): utilityai/llama-cpp-rs
Rust (automated build from crates.io): ShelbyJenkins/llm_client
C#/.NET: SciSharp/LLamaSharp
C#/VB.NET (more features - community license): LM-Kit.NET
Scala 3: donderom/llm4s
Clojure: phronmophobic/llama.clj
React Native: mybigday/llama.rn
Java: kherud/java-llama.cpp
Java: QuasarByte/llama-cpp-jna
Zig: deins/llama.cpp.zig
Flutter/Dart: netdur/llama_cpp_dart
Flutter: xuegao-tzx/Fllama
PHP (API bindings and features built on top of llama.cpp): distantmagic/resonance (more info)
Guile Scheme: guile_llama_cpp
Swift srgtuszy/llama-cpp-swift
Swift ShenghaiWang/SwiftLlama
Delphi Embarcadero/llama-cpp-delphi
Go (no CGo needed): hybridgroup/yzma
Android: llama.android

UIs

(to have a project listed here, it should clearly state that it depends on llama.cpp)

AI Sublime Text plugin (MIT)
BonzAI App (proprietary)
cztomsik/ava (MIT)
Dot (GPL)
eva (MIT)
iohub/collama (Apache-2.0)
janhq/jan (AGPL)
johnbean393/Sidekick (MIT)
KanTV (Apache-2.0)
KodiBot (GPL)
llama.vim (MIT)
LARS (AGPL)
Llama Assistant (GPL)
LlamaLib (Apache-2.0)
LLMFarm (MIT)
LLMUnity (MIT)
LMStudio (proprietary)
LocalAI (MIT)
LostRuins/koboldcpp (AGPL)
MindMac (proprietary)
MindWorkAI/AI-Studio (FSL-1.1-MIT)
Mobile-Artificial-Intelligence/maid (MIT)
Mozilla-Ocho/llamafile (Apache-2.0)
nat/openplayground (MIT)
nomic-ai/gpt4all (MIT)
ollama/ollama (MIT)
oobabooga/text-generation-webui (AGPL)
PocketPal AI (MIT)
psugihara/FreeChat (MIT)
ptsochantaris/emeltal (MIT)
pythops/tenere (AGPL)
ramalama (MIT)
semperai/amica (MIT)
withcatai/catai (MIT)
Autopen (GPL)

Tools

akx/ggify – download PyTorch models from HuggingFace Hub and convert them to GGML
akx/ollama-dl – download models from the Ollama library to be used directly with llama.cpp
crashr/gppm – launch llama.cpp instances utilizing NVIDIA Tesla P40 or P100 GPUs with reduced idle power consumption
gpustack/gguf-parser - review/check the GGUF file and estimate the memory usage
Styled Lines (proprietary licensed, async wrapper of inference part for game development in Unity3d with pre-built Mobile and Web platform wrappers and a model example)
unslothai/unsloth – 🦥 exports/saves fine-tuned and trained models to GGUF (Apache-2.0)

Infrastructure

Paddler - Open-source LLMOps platform for hosting and scaling AI in your own infrastructure
GPUStack - Manage GPU clusters for running LLMs
llama_cpp_canister - llama.cpp as a smart contract on the Internet Computer, using WebAssembly
llama-swap - transparent proxy that adds automatic model switching with llama-server
Kalavai - Crowdsource end to end LLM deployment at any scale
llmaz - ☸️ Easy, advanced inference platform for large language models on Kubernetes.
LLMKube - Kubernetes operator for llama.cpp with multi-GPU and Apple Silicon Metal support"

Games

Lucy's Labyrinth - A simple maze game where agents controlled by an AI model will try to trick you.

Supported backends

Backend	Target devices
Metal	Apple Silicon
BLAS	All
BLIS	All
SYCL	Intel and Nvidia GPU
OpenVINO [In Progress]	Intel CPUs, GPUs, and NPUs
MUSA	Moore Threads GPU
CUDA	Nvidia GPU
HIP	AMD GPU
ZenDNN	AMD CPU
Vulkan	GPU
CANN	Ascend NPU
OpenCL	Adreno GPU
IBM zDNN	IBM Z & LinuxONE
WebGPU [In Progress]	All
RPC	All
Hexagon [In Progress]	Snapdragon
VirtGPU	VirtGPU APIR

Obtaining and quantizing models

The Hugging Face platform hosts a number of LLMs compatible with llama.cpp:

You can either manually download the GGUF file or directly use any llama.cpp-compatible models from Hugging Face or other model hosting sites, such as ModelScope, by using this CLI argument: -hf <user>/<model>[:quant]. For example:

llama-cli -hf ggml-org/gemma-3-1b-it-GGUF

By default, the CLI would download from Hugging Face, you can switch to other options with the environment variable MODEL_ENDPOINT. For example, you may opt to downloading model checkpoints from ModelScope or other model sharing communities by setting the environment variable, e.g. MODEL_ENDPOINT=https://www.modelscope.cn/.

After downloading a model, use the CLI tools to run it locally - see below.

llama.cpp requires the model to be stored in the GGUF file format. Models in other data formats can be converted to GGUF using the convert_*.py Python scripts in this repo.

The Hugging Face platform provides a variety of online tools for converting, quantizing and hosting models with llama.cpp:

Use the GGUF-my-repo space to convert to GGUF format and quantize model weights to smaller sizes
Use the GGUF-my-LoRA space to convert LoRA adapters to GGUF format (more info: ggml-org#10123)
Use the GGUF-editor space to edit GGUF meta data in the browser (more info: ggml-org#9268)
Use the Inference Endpoints to directly host llama.cpp in the cloud (more info: ggml-org#9669)

To learn more about model quantization, read this documentation

`llama-cli`

A CLI tool for accessing and experimenting with most of `llama.cpp`'s functionality.

Run in conversation mode

Models with a built-in chat template will automatically activate conversation mode. If this doesn't occur, you can manually enable it by adding -cnv and specifying a suitable chat template with --chat-template NAME

llama-cli -m model.gguf

# > hi, who are you?
# Hi there! I'm your helpful assistant! I'm an AI-powered chatbot designed to assist and provide information to users like you. I'm here to help answer your questions, provide guidance, and offer support on a wide range of topics. I'm a friendly and knowledgeable AI, and I'm always happy to help with anything you need. What's on your mind, and how can I assist you today?
#
# > what is 1+1?
# Easy peasy! The answer to 1+1 is... 2!

Run in conversation mode with custom chat template

# use the "chatml" template (use -h to see the list of supported templates)
llama-cli -m model.gguf -cnv --chat-template chatml

# use a custom template
llama-cli -m model.gguf -cnv --in-prefix 'User: ' --reverse-prompt 'User:'

Constrain the output with a custom grammar
```
llama-cli -m model.gguf -n 256 --grammar-file grammars/json.gbnf -p 'Request: schedule a call at 8pm; Command:'

# {"appointmentTime": "8pm", "appointmentDetails": "schedule a a call"}
```
The grammars/ folder contains a handful of sample grammars. To write your own, check out the GBNF Guide.

For authoring more complex JSON grammars, check out https://grammar.intrinsiclabs.ai/

`llama-server`

A lightweight, OpenAI API compatible, HTTP server for serving LLMs.

Start a local HTTP server with default configuration on port 8080

llama-server -m model.gguf --port 8080

# Basic web UI can be accessed via browser: http://localhost:8080
# Chat completion endpoint: http://localhost:8080/v1/chat/completions

Support multiple-users and parallel decoding

# up to 4 concurrent requests, each with 4096 max context
llama-server -m model.gguf -c 16384 -np 4

Enable speculative decoding

# the draft.gguf model should be a small variant of the target model.gguf
llama-server -m model.gguf -md draft.gguf

Serve an embedding model

# use the /embedding endpoint
llama-server -m model.gguf --embedding --pooling cls -ub 8192

Serve a reranking model

# use the /reranking endpoint
llama-server -m model.gguf --reranking

Constrain all outputs with a grammar

# custom grammar
llama-server -m model.gguf --grammar-file grammar.gbnf

# JSON
llama-server -m model.gguf --grammar-file grammars/json.gbnf

`llama-perplexity`

A tool for measuring the perplexity ¹ (and other quality metrics) of a model over a given text.

Measure the perplexity over a text file

llama-perplexity -m model.gguf -f file.txt

# [1]15.2701,[2]5.4007,[3]5.3073,[4]6.2965,[5]5.8940,[6]5.6096,[7]5.7942,[8]4.9297, ...
# Final estimate: PPL = 5.4007 +/- 0.67339

Measure KL divergence
```
# TODO
```

`llama-bench`

Benchmark the performance of the inference for various parameters.

Run default benchmark

llama-bench -m model.gguf

# Output:
# | model               |       size |     params | backend    | threads |          test |                  t/s |
# | ------------------- | ---------: | ---------: | ---------- | ------: | ------------: | -------------------: |
# | qwen2 1.5B Q4_0     | 885.97 MiB |     1.54 B | Metal,BLAS |      16 |         pp512 |      5765.41 ± 20.55 |
# | qwen2 1.5B Q4_0     | 885.97 MiB |     1.54 B | Metal,BLAS |      16 |         tg128 |        197.71 ± 0.81 |
#
# build: 3e0ba0e60 (4229)

`llama-simple`

A minimal example for implementing apps with `llama.cpp`. Useful for developers.

Basic text completion

llama-simple -m model.gguf

# Hello my name is Kaitlyn and I am a 16 year old girl. I am a junior in high school and I am currently taking a class called "The Art of

Contributing

Contributors can open PRs
Collaborators will be invited based on contributions
Maintainers can push to branches in the llama.cpp repo and merge PRs into the master branch
Any help with managing issues, PRs and projects is very appreciated!
See good first issues for tasks suitable for first contributions
Read the CONTRIBUTING.md for more information
Make sure to read this: Inference at the edge
A bit of backstory for those who are interested: Changelog podcast

XCFramework

The XCFramework is a precompiled version of the library for iOS, visionOS, tvOS, and macOS. It can be used in Swift projects without the need to compile the library from source. For example:

// swift-tools-version: 5.10
// The swift-tools-version declares the minimum version of Swift required to build this package.

import PackageDescription

let package = Package(
    name: "MyLlamaPackage",
    targets: [
        .executableTarget(
            name: "MyLlamaPackage",
            dependencies: [
                "LlamaFramework"
            ]),
        .binaryTarget(
            name: "LlamaFramework",
            url: "https://github.com/ggml-org/llama.cpp/releases/download/b5046/llama-b5046-xcframework.zip",
            checksum: "c19be78b5f00d8d29a25da41042cb7afa094cbf6280a225abe614b03b20029ab"
        )
    ]
)

The above example is using an intermediate build b5046 of the library. This can be modified to use a different version by changing the URL and checksum.

Completions

Command-line completion is available for some environments.

Bash Completion

$ build/bin/llama-cli --completion-bash > ~/.llama-completion.bash
$ source ~/.llama-completion.bash

Optionally this can be added to your .bashrc or .bash_profile to load it automatically. For example:

$ echo "source ~/.llama-completion.bash" >> ~/.bashrc

Dependencies

yhirose/cpp-httplib - Single-header HTTP server, used by llama-server - MIT license
stb-image - Single-header image format decoder, used by multimodal subsystem - Public domain
nlohmann/json - Single-header JSON library, used by various tools/examples - MIT License
miniaudio.h - Single-header audio format decoder, used by multimodal subsystem - Public domain
subprocess.h - Single-header process launching solution for C and C++ - Public domain

https://huggingface.co/docs/transformers/perplexity ↩

Name		Name	Last commit message	Last commit date
Latest commit History 8,545 Commits
.devops		.devops
.gemini		.gemini
.github		.github
benches		benches
ci		ci
cmake		cmake
common		common
docs		docs
examples		examples
ggml		ggml
gguf-py		gguf-py
grammars		grammars
include		include
licenses		licenses
media		media
models		models
pocs		pocs
requirements		requirements
scripts		scripts
src		src
tests		tests
tools		tools
vendor		vendor
.clang-format		.clang-format
.clang-tidy		.clang-tidy
.dockerignore		.dockerignore
.ecrc		.ecrc
.editorconfig		.editorconfig
.flake8		.flake8
.gitignore		.gitignore
.gitmodules		.gitmodules
.pre-commit-config.yaml		.pre-commit-config.yaml
AGENTS.md		AGENTS.md
AUTHORS		AUTHORS
CLAUDE.md		CLAUDE.md
CMakeLists.txt		CMakeLists.txt
CMakePresets.json		CMakePresets.json
CODEOWNERS		CODEOWNERS
CONTRIBUTING.md		CONTRIBUTING.md
LICENSE		LICENSE
Makefile		Makefile
README.md		README.md
SECURITY.md		SECURITY.md
TURBOQUANT_CUDA_IMPLEMENTATION.md		TURBOQUANT_CUDA_IMPLEMENTATION.md
apply_turbo_cuda_v2.py		apply_turbo_cuda_v2.py
benchmark-results.md		benchmark-results.md
build-xcframework.sh		build-xcframework.sh
convert_hf_to_gguf.py		convert_hf_to_gguf.py
convert_hf_to_gguf_update.py		convert_hf_to_gguf_update.py
convert_llama_ggml_to_gguf.py		convert_llama_ggml_to_gguf.py
convert_lora_to_gguf.py		convert_lora_to_gguf.py
experiments.md		experiments.md
flake.lock		flake.lock
flake.nix		flake.nix
github1.png		github1.png
github2.png		github2.png
mypy.ini		mypy.ini
notes-for-tom.md		notes-for-tom.md
poetry.lock		poetry.lock
pyproject.toml		pyproject.toml
pyrightconfig.json		pyrightconfig.json
requirements.txt		requirements.txt

Folders and files

Latest commit

History

Repository files navigation

llama.cpp

TurboQuant CUDA KV Cache Quantization

What's implemented

Benchmarks (Qwen3.5-27B Q6_K, RTX 3090 24GB)

Quality (Perplexity, lower is better)

Speed

Extreme Context (128K on 24GB — impossible with q8_0)

Usage

Context length recommendations

Key findings

Recent API changes

Hot topics

Quick start

Description

Text-only

Multimodal

Supported backends

Obtaining and quantizing models

A CLI tool for accessing and experimenting with most of llama.cpp's functionality.

A lightweight, OpenAI API compatible, HTTP server for serving LLMs.

A tool for measuring the perplexity 1 (and other quality metrics) of a model over a given text.

Benchmark the performance of the inference for various parameters.

A minimal example for implementing apps with llama.cpp. Useful for developers.

Contributing

Other documentation

Development documentation

Seminal papers and background on the models

XCFramework

Completions

Bash Completion

Dependencies

Footnotes

About

Resources

License

Contributing

Security policy

Uh oh!

Stars

Watchers

Forks

Uh oh!

Uh oh!

Languages

A CLI tool for accessing and experimenting with most of `llama.cpp`'s functionality.

A tool for measuring the perplexity ¹ (and other quality metrics) of a model over a given text.

A minimal example for implementing apps with `llama.cpp`. Useful for developers.