Ollama: run AI models on your own computer
Open-source project on GitHub that lets you download and use open AI models offline, on your computer. Explained without jargon.
What it is
Ollama is an open-source project hosted on GitHub. It lets you download open-weight AI models and use them on your own computer: Gemma (Google), Qwen, DeepSeek or gpt-oss (OpenAI), among others, according to its own repository.
Why it matters
When you use ChatGPT or Gemini, what you type travels to the company’s servers. With a local model, the conversation stays on your device. That can be an interesting option for personal texts.
What to keep in mind
- Local models are usually less capable than the big cloud assistants.
- They need memory and disk space: the largest ones won’t fit on an ordinary laptop.
- Only download from the official website or repository.
How to try it
Only download it from ollama.com (Windows, macOS or Linux) and follow its instructions to choose a model. Start with a small one.
Official links
Plain-language summary based on the official sources linked. Cognenza has no commercial relationship with the companies mentioned.