Skip to content

Releases: ArthurYangX/nano-NotebookLM

v0.2.0 — initial public release

22 May 17:18
@ArthurYangX ArthurYangX
v0.2.0
71bcb38

Choose a tag to compare

View all tags
v0.2.0 — initial public release Latest
Latest

nano-NotebookLM v0.2.0 — initial public release.

A self-hosted study assistant: upload course PDFs / PPTX / DOCX / Markdown, and chat with page-accurate citations, get LaTeX notes, exam-prep with a self-evolving question bank, and an editable knowledge graph — all running on the LLM of your choice.

Highlights

  • Bring-your-own-model: OpenAI · DeepSeek · Moonshot · Zhipu · MiniMax · Groq · Together · Anthropic Claude · Gemini, or any OpenAI-compatible local runner (Ollama / vLLM / LM Studio / llama.cpp). UI-managed provider matrix, swap without restart.
  • Page-accurate citations via BM25 + FAISS + RRF + knowledge-graph retrieval. Every answer links back to the source page in the built-in PDF reader.
  • Three-tier embeddings with hot-swap: local MiniLM (offline) / OpenAI text-embedding-3-large / BGE-M3.
  • MinerU OCR ingest for scanned/formula-heavy PDFs (optional extra; auto-detects CUDA on Linux).
  • Background upload pipeline with NDJSON stage events — close the tab, come back, ingest keeps running.
  • No DB, no auth, no build step: single FastAPI process + CDN React 18. Everything lives under ./artifacts/.

Quick start

git clone https://github.com/ArthurYangX/nano-NotebookLM && cd nano-NotebookLM
uv venv && source .venv/bin/activate && uv pip install -e ".[test]"
cp .env.example .env   # fill at least one LLM key
python api/server.py   # → http://localhost:8000

See README.md for full setup, optional extras (MinerU / tectonic / LibreOffice), and the provider matrix.

Not affiliated with Google. NotebookLM is a trademark of Google LLC.

Assets 2
Loading