// the find
pinecone-io/examples
Jupyter Notebooks to help you get hands-on with Pinecone vector databases
Pinecone's official example repository: Jupyter notebooks covering RAG, semantic search, recommendation systems, and various integrations with LangChain, LlamaIndex, and different LLMs. Aimed at developers evaluating or getting started with Pinecone who want working code rather than docs alone.
The split between `docs/` (reviewed, production-ready) and `learn/` (exploratory) is an honest signal about what you can trust in production. CI runs notebooks on every PR (`test-notebooks-changed.yaml`) which means most examples actually execute rather than just looking correct. Coverage of hybrid search, reranking, and chunking strategies goes beyond the usual 'embed and query' toy demos. The BERTopic, RAGAS evaluation, and offline evaluation metric notebooks are genuinely useful reference material independent of Pinecone.
Everything here assumes you're using Pinecone's managed service — there's no path for self-hosted or local vector DB alternatives, so the notebooks double as sales material. The `learn/` notebooks have no dependency pinning policy (version-census.py exists but isn't enforced), meaning older notebooks silently break as OpenAI and LangChain release breaking changes. Several notebooks in `learn/generation/llm-field-guide/` reference models (Falcon, MPT, GPT-J) that are now largely obsolete with no indication they've been deprecated. Structurally this repo is a flat dump of 150+ notebooks with minimal cross-linking, making it hard to find the right starting point without already knowing what you're looking for.