llama-cpp

A toolchain for running large language models locally with llama.cpp and the GGUF file format, which stores model weights in compact, often reduced-precision form.

In plain words
What is it for?
Use it to run local models, convert and reduce their size, serve them through an OpenAI-compatible interface, and deploy them on laptops, edge devices, or supported GPUs.
Why use it?
It lets you run and serve models on CPUs, Apple Silicon, and several GPU types without depending on a specific NVIDIA setup or a Python runtime.

Skill for Claude CodeCodex

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/moltis-org/moltis/llama-cpp
Any agent
npx skills add moltis-org/moltis --skill llama-cpp
Clone the repo
git clone --depth 1 https://github.com/moltis-org/moltis

Made for: Claude Code, Codex.

Per session 91 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,991 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 1 finding. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00091 $0.03991
Opus 5 $0.00046 $0.01996
Sonnet 5 $0.00018 $0.00798
Haiku 4.5 $0.00009 $0.00399

Measured 3d ago against content hash f27402ad8529, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

llama-cpp scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl http://localhost:8080/v1/chat/completions \
crates/skills/src/assets/mlops/inference/llama-cpp/SKILL.md · 436 lines

How it starts

The opening of the file, as written. The whole thing — 436 lines — stays where its author put it; the contents beside it link to each section on GitHub.

llama.cpp + GGUF

Pure C/C++ LLM inference with minimal dependencies, plus the GGUF (GPT-Generated Unified Format) standard used for quantized weights. One toolchain covers conversion, quantization, and serving.

When to use

Use llama.cpp + GGUF when:

  • Running on CPU-only machines or Apple Silicon (M1/M2/M3/M4) with Metal acceleration
  • Using AMD (ROCm) or Intel GPUs where CUDA isn't available
  • Edge deployment (Raspberry Pi, embedded systems, consumer laptops)
  • Need flexible quantization (2–8 bit with K-quants)
  • Want local AI tools (LM Studio, Ollama, text-generation-webui, koboldcpp)
  • Want a single binary deploy without Docker/Python

Key advantages:

  • Universal hardware: CPU, Apple Silicon, NVIDIA, AMD, Intel
  • No Python runtime required (pure C/C++)
  • K-quants + imatrix for better low-bit quality
  • OpenAI-compatible server built in
  • Rich ecosystem (Ollama, LM Studio, llama-cpp-python)

Use alternatives instead:

  • vLLM — NVIDIA GPUs, PagedAttention, Python-first, max throughput
  • TensorRT-LLM — Production NVIDIA (A100/H100), maximum speed
  • AWQ/GPTQ — Calibrated quantization for NVIDIA-only deployments
  • bitsandbytes — Simple HuggingFace transformers integration
  • HQQ — Fast calibration-free quantization

Quick start

Install

# macOS / Linux (simplest)
brew install llama.cpp

# Or build from source
git clone https://github.com/ggml-org/llama.cpp
cd llama.cpp
make                        # CPU
make GGML_METAL=1           # Apple Silicon
make GGML_CUDA=1            # NVIDIA CUDA
make LLAMA_HIP=1            # AMD ROCm

# Python bindings (optional)
pip install llama-cpp-python
# With CUDA:   CMAKE_ARGS="-DGGML_CUDA=on" pip install llama-cpp-python --force-reinstall --no-cache-dir
# With Metal:  CMAKE_ARGS="-DGGML_METAL=on" pip install llama-cpp-python --force-reinstall --no-cache-dir

Download a pre-quantized GGUF

# TheBloke hosts most popular models pre-quantized
huggingface-cli download \
    TheBloke/Llama-2-7B-Chat-GGUF \
    llama-2-7b-chat.Q4_K_M.gguf \
    --local-dir models/

Read the full file on GitHub · 436 lines

Files

What ships with it

5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago First seen · 436 lines · 91 tokens per session scan A f27402ad8529

Subscribe to this mod's changes

llama-cpp is a skill published in the GitHub repository moltis-org/moltis (2,841 stars, last pushed today), licensed MIT. It adds 91 tokens to every session and 3,991 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.