Skip to content
ToolOpen-source agent

ai-agents-from-scratch

Step-by-step tutorial for building AI agents locally in JavaScript with node-llama-cpp

Pricing
Open source
Platforms
self-hosted
GitHub stars
4.9K
Last commit
4 months ago
Language
JavaScript
License
MIT

About ai-agents-from-scratch

ai-agents-from-scratch is an open-source AI tool for ai models and llm tooling. It is available on self-hosted.

Instead of reaching for a framework, the course walks through each layer by hand: loading a local model, the prompt and response cycle, tool use, and finally a ReAct-style agent. Each lesson has runnable code plus separate code explanations and concept notes.

Examples run with Node.js 18 or newer and local models placed in a models folder; the README suggests at least 8 GB of RAM and 16 GB if possible. An optional lesson shows how to call a hosted model, and a Python version lives in a sibling repository. A companion website explains the learning path visually.

What it does

Local-model examples
Lessons use node-llama-cpp so no hosted API is needed.
Progressive learning path
Starts with a basic LLM call and builds to a ReAct agent.
Code and concept notes
Each example links to a code walkthrough and a concepts page.
Optional hosted lesson
One lesson covers calling a hosted model, temperature and token usage.
Companion website
A site maps the learning path and explains each example.

Good for

  • Learn how agents work before adopting a framework
  • Run an LLM on your own machine from Node.js
  • Study ReAct agent structure with working code
  • Teach agent fundamentals in a workshop

Pricing

Open source

MIT licensed tutorial

ai-agents-from-scratch FAQ

What do I need to run it?

Node.js 18 or newer and at least 8 GB of RAM, with model files downloaded into the models folder.

Is there a Python version?

Yes, the README links to a separate Python tutorial repository.

What is the licence?

MIT.

Is ai-agents-from-scratch free?

Yes. ai-agents-from-scratch is open source, so you can use and self-host it for free.

What are the best alternatives to ai-agents-from-scratch?

The closest alternatives to ai-agents-from-scratch are ECC, ponytail, graphify, ragflow and Agent-Reach.

Alternatives to ai-agents-from-scratch

  1. Skills, hooks, memory and security layer for coding agent harnesses

    • Freemium
    • Open-source agent
    • 305K GitHub stars
    • AI models and LLM tooling
  2. Ruleset that makes AI coding agents write less code that still works

    • Open source
    • Open-source agent
    • 192K GitHub stars
    • AI models and LLM tooling
  3. Code knowledge graph that gives coding assistants shared, source-linked context

    • Freemium
    • Open-source agent
    • 149K GitHub stars
    • Knowledge and memory
  4. Open-source RAG engine with a document pipeline, hybrid search and visual agent workflows

    • Freemium
    • Open-source agent
    • 94.3K GitHub stars
    • Knowledge and memory
  5. Command-line channels that let AI agents read X, Reddit, YouTube and Chinese platforms

    • Open source
    • Open-source agent
    • 112K GitHub stars
    • AI models and LLM tooling
  6. Self-hosted AI gateway routing many providers through one local endpoint with fallback

    • Freemium
    • Open-source agent
    • 88.1K GitHub stars
    • AI models and LLM tooling