Running LLMs on Your Own Machine

A nine-minute tour of running large language models locally — no cloud, no API keys, no monthly bill.

📽 Watch the presentation

→ caseproof.github.io/running-llms-locally

What you'll learn

• Why local — privacy, cost, offline access, and control
• The three tools to know — GPT4All, LM Studio, and Ollama, and which one fits you
• Matching models to your hardware — how much RAM you need for 3B, 7B, 13B, and 70B models
• LocalDocs & RAG — turn any folder into searchable, cited context
• The Ollama API — point any OpenAI-compatible app at your laptop in two lines
• Modelfiles — bake custom personas into a model and share them like a gist
• Obsidian integration — make your notes the AI's long-term memory
• Mobile — the best on-device LLM apps for iOS and Android