AI Integration
Ollama Tutorial for Snipset
Set up Ollama locally and connect it to Snipset: install, model selection, hardware sizing, semantic search, snippet generation, and the AI Hub.
Ollama is the easiest way to run Large Language Models (LLMs) like Gemma, Llama 3, or DeepSeek, either locally on your own machine or through Ollama Cloud. This guide covers how to quickly start interactive chats and manage your models.
Part 1: Installation
If you haven't installed Ollama yet, please follow our dedicated setup guides based on your preference:
- Beginner Installation (Recommended): Visit our Getting Started with Ollama guide for the easiest and fastest way to install the standard version using the official
.exeinstaller. - Power User Installation: If you want to optimize storage and run a custom CPU-only setup, read the Advanced Power User Guide.
Part 2: Quickstart Chat
Once Ollama is installed, let's run your first AI model. Open PowerShell or Command Prompt. Cloud models need a one-time sign-in, so run ollama signin first, then type the following command:
`ollama run gemma4:31b-cloud`
Part 3: Model Management via CLI
Ollama operates via a Command Line Interface (CLI). Here are the essential commands every user should know:
| Command | Function | Example |
|---|---|---|
| `run` | Runs the model (and downloads it if necessary). | `ollama run gemma4:31b-cloud` |
| `pull` | Downloads a model to your machine without running an interactive chat. | `ollama pull nomic-embed-text` |
| `list` | Shows a list of all models currently installed on your computer. | `ollama list` |
| `rm` | Removes a model you no longer use to free up disk space. | `ollama rm mistral` |
| `ps` | Displays which models are currently loaded and running in your memory. | `ollama ps` |
| `show` | Displays technical information about a model (like parameters and licenses). | `ollama show gemma4:31b-cloud` |
Part 4: Snipset Integration
Once Ollama is installed and a chat model has been pulled (we recommend gemma4:31b-cloud), Snipset will automatically detect your local Ollama service running at http://localhost:11434. Open Settings → AI & Models → AI Configuration, select the model from the list, and you're ready to use AI Chat and automated snippet generation. Cloud models need internet; for fully offline use, pick a local model such as llama3.2:3b.