Tiempo de lectura: 2 minutosMCP (Model Context Protocol) is an open standard, created by Anthropic, that defines how an AI assistant connects to external tools and data. It is the common way to say to a model: «here are these functions you can use». The MCPs are defined in the opencode.json file, under the key mcp. You have two … Read more
Tiempo de lectura: 2 minutosOllama does not support native tags from some Hugging Face repositories (such as I-Compact, Balanced, etc.) when they do not conform to their standard quantization schema (Q4_K_M, Q5_K_S…). The solution is to manually download the .gguf and register it with a Modelfile. Repo: https://huggingface.co/mudler/Qwen3.6-35B-A3B-APEX-GGUF For a 16GB RTX 5060 Ti, the suitable variant is I-Compact … Read more
Tiempo de lectura: < 1 minutoToday we are going to increase the response time allowed in an AI local model on Ollama that functions below the security perimeter using a Nginx Proxy Manager. To do this, we are going to Nginx Proxy Manager. We open Proxy Hosts > We open our host > Select Edit and we go to Advanced. … Read more
Tiempo de lectura: 2 minutosI’ll show you a comparison of the best local LLM models for programming. The entire benchmark was performed on a server with the following features: OLLAMA_KEEP_ALIVE=-1 OLLAMA_KV_CACHE_TYPE=q4_0 OLLAMA_FLASH_ATTENTION=true OLLAMA_NEW_ENGINE=true Methodology To compare models uniformly, the same prompt was used in all cases: “Write a Python class called GestorTareas with methods: add task, delete task, mark … Read more
Tiempo de lectura: 2 minutosToday we will explain how to use Ollama from a custom provider for GitHub Copilot. The first thing we need to have is GitHub Copilot installed in our VS Code And also Ollama installed (recommended in a remote environment). Once all the necessary services are ready, we proceed with action. We open VS Code and … Read more
Tiempo de lectura: < 1 minutoHoy vamos a aprendir a instalar OpenClaw desde cero en una Raspberry Pi y con una tarjeta SD. Do not use the normal Windows or Mac formatter. Use Raspberry Pi Imager (the official tool). No update until you plug the SD into the Pi and enter via SSH: Bash sudo apt update && sudo apt … Read more
Tiempo de lectura: < 1 minutoHoy vamos a indicar cuales son los modelos Ollama Cloud que no dan el error de: {«error»:»this model requires a subscription, upgrade for access: https://ollama.com/upgrade (ref: 374c814e-dc9c-4fc1-a898-0bbc57cb0d0a)»} To determine this and by running tests with this script we can know which models are available to use with cloud and free tier cat
Tiempo de lectura: < 1 minutoHoy vamos a aprender a instalar OpenClaw en un sistema Docker mediante Docker Compose y dejarlo de forma aislada. Primero instalamos Docker. Once installed, we need to update Docker Compose: # Add official Docker repository curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo gpg –dearmor -o /usr/share/keyrings/docker-archive-keyring.gpg echo “deb [arch=$(dpkg –print-architecture) signed-by=/usr/share/keyrings/docker-archive-keyring.gpg] https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable” | sudo … Read more
Tiempo de lectura: < 1 minutoThis is the day we will learn how to create a Google Gemini IA client for Python in this way, so we can use the Gemini model from Google for our projects in Python. The first thing we need to do is install this library: pip install google-generativeai pip install google-genai You will see how … Read more
Tiempo de lectura: 2 minutosWe are going to install OpenClaw on a Raspberry Pi 4 and use it as a bot controlled by Telegram today. First on the Pi, update the system and install the basic dependencies. sudo apt update && sudo apt upgrade -y sudo apt install -y git curl build-essential sudo timedatectl set-timezone Europe/Madrid Install Node.js 24. … Read more
