With the CodeSky desktop app, you can run AI models right on your own computer using Ollama. This gives you a private, offline way to build, all without sending your work to the cloud.
What are local AI models?
Normally, the AI that powers CodeSky runs in the cloud. Local AI models are different: the model runs on your own computer instead. CodeSky supports this through Ollama, the named feature for running models locally.
The result is the same conversational way of building you already know, but powered by a model that lives on your machine.
Why use local models?
- Privacy. Your prompts and your project stay on your own computer, which is reassuring for sensitive or confidential ideas.
- Offline use. Because the model runs locally, you can keep building even when you do not have an internet connection.
- Independence. You are not relying on a constant connection to a cloud service to make progress.
What you need
Local AI models are available through the desktop app, since the model runs on your computer. To use them, you will have Ollama set up on your machine, and then choose a local model inside CodeSky when you build.
How it works for you
- Open the desktop app on your computer.
- Choose a local model as the one that will power your building.
- Describe what you want, just like always, and let the local model do the work.
Everything else feels familiar. You still describe your idea in plain language, preview the result, and refine with follow-up requests.
Things to keep in mind
- Your computer does the work. Local models use your own machine, so a more capable computer generally gives a smoother experience.
- Different models, different strengths. Just like cloud models, local ones vary, so you can pick the one that fits your project.
- Cloud is still there. You can switch back to cloud models any time you prefer.
Local AI models are a great fit when privacy and offline freedom matter most, giving you a self-contained way to build entirely on your own terms.