Skip to content
Local Model Notes Where to start

Site map

  1. What running an AI model locally means

    What changes when an AI language model runs on your own computer instead of a web chatbot, and what you'd actually need to download and keep.

  2. What computer hardware a local model needs

    Which parts of a laptop or office PC matter for running an AI model locally, with the memory and chip figures two free tools publish.

  3. Model sizes and memory explained

    What 7B, 70B and similar labels mean for an AI model, and a simple way to estimate how much memory a downloaded model needs.

  4. Quantization in plain English

    Why one AI model is offered in several file sizes, what quantization does to the numbers inside it, and what you give up for a smaller file.

  5. How Ollama works

    What Ollama does, how it downloads and runs a model on your own computer, and what its own documentation says about memory, storage and cloud features.

  6. How LM Studio works

    What LM Studio is, how its desktop app finds, downloads and runs models, and how it differs from command-line tools, based on its own documentation.

  7. Open-weight model licences explained

    What "open weight" means, how it differs from open source, and what two real licences say about business use. A guide to reading terms, not legal advice.

  8. What stays on your computer and what does not

    What a local AI model keeps on your machine, and which jobs in Ollama and LM Studio still use the internet or can send data elsewhere.