Skip to content
Back to AI Tools

Ollama

Categories
  • Local & Open Source
Pricing
Free / Freemium
Visit Website

Ollama is the fastest and simplest way to run large language models locally. With a single command (`ollama run llama3.3`), you download and start chatting with a model. It provides a clean REST API that is compatible with countless tools — OpenCode, Cherry Studio, Continue.dev, LangChain, and more can all connect to Ollama as their backend. For developers and researchers who want local AI as infrastructure rather than a standalone app, Ollama's lightweight server model and extensive integration ecosystem make it the default choice.

Key Features

  • One-command model download and execution: `ollama run <model>`
  • REST API for programmatic access from any language or tool
  • Model customization via Modelfile — adjust parameters, system prompts, and more
  • Broad ecosystem integration: OpenCode, Cherry Studio, Continue, LangChain, LlamaIndex
  • Cross-platform: macOS, Linux, Windows with GPU acceleration

Usage Guide

Setting Up Ollama

  1. Download and install from ollama.com, or use `brew install ollama` on macOS.
  2. Pull a model: `ollama pull llama3.3` or `ollama pull deepseek-r1:8b`.
  3. Start chatting: `ollama run llama3.3` and type your prompts.
  4. Access via API: send requests to `http://localhost:11434/api/generate`.
  5. Connect tools: point OpenCode, Cherry Studio, or Continue to `http://localhost:11434` as the model provider.

What can it achieve?

  • Development Backend

    Run Ollama as a local API server and connect coding tools like OpenCode or Cherry Studio for fully offline development.

  • Research Infrastructure

    Provide a shared local LLM service for a lab group — everyone uses the same models with consistent configuration.

  • Custom Model Experiments

    Test fine-tuned or custom models locally before deploying to production, with full control over parameters and prompts.