Aulafy
Courses/Local MLOps and model deployment/llama.cpp server locally

llama.cpp server locally

llama.cpp server is a straightforward way to serve GGUF models locally. It's lightweight, controllable, and perfect for learning what it means to have a model behind your own API.

  • Build or install llama.cpp and start a local server.
  • Test an HTTP request without relying on a graphical interface.
  • Record the model, port, context, and parameters used.

Minimal server

TerminalCode
git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build --config Release

./build/bin/llama-server \
  -m models/modelo.gguf \
  --host 127.0.0.1 \
  --port 8080 \
  -c 4096

Test with curl

TerminalCode
curl http://127.0.0.1:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "local-gguf",
    "messages": [
      {"role": "user", "content": "Explica qué es una cola en MLOps IA en 3 frases"}
    ],
    "temperature": 0.2
  }'

Service manifest

TerminalCode
servicio: llama-server-local
modelo: models/modelo.gguf
hash_modelo: sha256:...
host: 127.0.0.1
puerto: 8080
contexto: 4096
cuantizacion: Q4_K_M
uso: desarrollo local
fecha: 2026-07-03