Free Ollama Update Makes Running AI on Your Own Laptop Smoother

Ollama is a free, open-source app that lets you run AI models — the same kind of technology behind ChatGPT — on your own laptop or desktop instead of on a company's servers. It has become hugely popular: more than 182,000 people have "starred" it on GitHub, the code-sharing site. On 23 September, the team shipped version 0.34.4, a maintenance release that fixes bugs rather than adding flashy new features.

The headline fix targets a frustrating glitch: intermittent "model not found" errors, where an AI you had already downloaded would mysteriously disappear mid-conversation. The update also makes Qwen models (a widely used family of AI assistants) process your prompts faster, and lets Gemma image models automatically pick the right picture resolution instead of guessing. There's also a small tweak to how the Mac app detects other AI apps you may already have installed.

Why should you care? Because Ollama is one of the simplest ways to try a genuinely private AI. Everything runs on your own hardware, so your notes, documents, medical questions or embarrassing searches never leave your computer — and you never pay a monthly subscription. For anyone uneasy about sending everything they type to a giant tech company, tools like this are the alternative. Small improvements like bug fixes matter here: reliability is the difference between a fun weekend toy and something you actually use every day.

The catch: this is a "pre-release," meaning it's a test build that may still have rough edges. Local AI also demands a reasonably modern computer with plenty of memory, and the free models you can run at home are generally less capable than the best paid cloud versions — think competent assistant rather than genius. Still, if you've been curious about AI that works offline, this is a good week to try it.