What software do I need to actually use these models?
You don't need to write code or deal with complicated setup to run local AI. The most popular tools are free and run across Windows, Mac, and Linux. Which one makes sense comes down to how you want to use it:
- All-in-One Desktop Apps (Best for Beginners): Tools like LM Studio and Jan give you a familiar, ChatGPT-style window. Open the app, search the built-in catalog, click Download, and start chatting. LM Studio even checks your hardware first and warns you if a model won't fit in your computer's memory before you bother downloading it.
- Document & Knowledge-Base Apps: If you want to ask questions against your own PDFs, notes, or local folders without uploading them anywhere, apps like AnythingLLM or GPT4All pair a local AI runner with built-in document search.
- Developer Engines & Web Interfaces: Ollama runs quietly in the background as a lightweight local engine. Developers use it to wire local models into editors like VS Code or Cursor, or pair it with Open WebUI for a ChatGPT-style interface in the browser.
Under the hood, almost all of these apps rely on the same open-source inference engine (llama.cpp), so a model will generate text at roughly the same speed no matter which one you pick. Nothing stops you from installing both LM Studio and Ollama on the same machine to see which workflow sticks.
Looking for verified hardware rigs?
Check our real-time £/GB VRAM index to find systems optimized for local model inference.