README.md

---
title: OpenKB Llama Tutorial — Marktechpost
emoji: 📚
colorFrom: indigo
colorTo: blue
sdk: gradio
sdk_version: "4.44.0"
app_file: app.py
pinned: false
license: apache-2.0
short_description: AI wiki from docs via OpenKB + Llama 3.3-70B
---

# OpenKB + OpenRouter (Llama 3.3-70B) — Knowledge Base Tutorial

**by [Marktechpost](https://marktechpost.com)**

Build an AI-powered knowledge wiki from research documents using
[OpenKB](https://github.com/VectifyAI/OpenKB) and Meta's **Llama 3.3-70B Instruct**
served for free via [OpenRouter](https://openrouter.ai).

## What this demo covers

| Step | Description |
|------|-------------|
| 1 | Initialise a local knowledge base |
| 2 | Compile 3 AI research docs (Transformers, RAG, Knowledge Graphs) into a wiki |
| 3 | Explore auto-generated summaries and concept pages |
| 4 | `openkb list` + `openkb status` |
| 5 | Natural-language queries against the wiki |
| 6 | Save a deep synthesis query |
| 7 | Lint the wiki for health issues |
| 8 | Programmatic cross-reference graph analysis |
| 9 | Incremental update — add a 4th document (Sparse Attention) |

## Prerequisites

1. Create a **free** account at [openrouter.ai](https://openrouter.ai).
2. Generate an API key at [openrouter.ai/keys](https://openrouter.ai/keys).
3. Paste the key into the app — it is used only for your session and never stored.

The model used (`meta-llama/llama-3.3-70b-instruct:free`) is **free with no credit card required**.

## Other free models to try

```
openrouter/mistralai/mistral-7b-instruct:free
openrouter/google/gemma-3-27b-it:free
openrouter/qwen/qwen3-14b:free
openrouter/microsoft/phi-4-reasoning:free
```

Change `LLM_MODEL` in `app.py` to switch models.

## Local development

```bash
pip install -r requirements.txt
python app.py
```

## Credits

- [OpenKB](https://github.com/VectifyAI/OpenKB) by VectifyAI
- Tutorial by [Marktechpost](https://marktechpost.com)