Hermes Agent + Ollama
Your 100% private AI OS
Run a complete AI agent on your own machine — no monthly fee, no rate limit, no internet, and without any data leaving your home. From the fundamentals to 7 step-by-step projects.
0 of 114 topics read · mark each section as you go
What you’ll build
This course takes the concept of Local, private AI and takes you from theory to practice: installing Ollama, choose and download open models (Qwen, DeepSeek, Gemma, Mistral), connect a model to the Hermes Agent and build your own "AI operating system"—everything running on your machine, for free and forever.
The core idea is ownership: stop renting intelligence. Your data never leaves home, you don't pay per token, it works without internet (on a plane, off-grid), and there's no gatekeeper. Where the cloud still wins in raw quality, you learn to switch — without being ideological, using the right tool for each task.
100% private
The data never leaves your machine.
$0 per token
Download once, run forever.
Works offline
No internet, no rate limits.
You’re in charge
No vendor lock-in, no gatekeeper.
The 3 tracks
🧠 Fundamentals
Why local is the future, the vocabulary (LLM, agent, AI OS), Ollama, context window, and the three modes.
🛠️ Hands-on
Install Ollama, choose the model for your hardware, download/chat, set up Qwen 3 Coder 64k, and connect to Hermes.
🚀 Projects
7 projects step by step: private chat, local agent, memory/personas/skills, GitHub, 24/7 background agents, and mobile.
How the course works
Mark what you've read
Each section has a “mark as read” button. Progress by module, track, and course is calculated automatically.
Take notes right in the text
Select a passage to highlight, annotate, or mark as a question. Everything is saved in your browser.
My journey
A dashboard brings together progress, questions, and notes, and takes you back to where you left off. Export/import in .json.
Copy-and-run examples
Real Ollama commands, ready to paste and run, with the expected results.
Real screenshots
The screens come straight from the source video: Ollama, terminal, the Hermes app, model selection, and benchmarks.
Read your way
Themes (dark/light/sepia/focus/high contrast), font size, width, and line spacing in the “Appearance” button.
✅ Prerequisites
- •A computer (Mac, Linux, or Windows) with some free disk space—models take up ~2 GB to dozens of GB.
- •Willingness to open the terminal a few times (we explain every command).
- •No prior AI knowledge required—Track 1 defines every term from scratch.