// the find
weaviate/recipes
This repository shares end-to-end notebooks on how to use various Weaviate features and integrations!
A collection of Jupyter notebooks showing how to use Weaviate with various tools, frameworks, and cloud providers. It's the official example repo for Weaviate's integrations — useful if you're evaluating Weaviate or trying to connect it to something specific like LangChain, Databricks, or Confluent.
The breadth of integrations is genuinely useful — covering LLM frameworks (LangChain, LlamaIndex, DSPy, CrewAI), data platforms (Spark, Airbyte, Databricks), and cloud providers (AWS, GCP, Azure) in one place saves significant search time. The notebooks include working end-to-end examples rather than just API snippets, so you can actually run them to verify something works before committing. The multi-vector ColBERT and Product Quantization examples are worth attention if you care about search quality vs. memory tradeoffs — these aren't topics most vector DB example repos bother to demonstrate. The contributor tooling to convert notebooks to docs-friendly markdown is a small but practical detail that helps keep examples in sync with official docs.
The repo is vendor-specific marketing material dressed as neutral examples — every notebook naturally concludes that Weaviate is the right choice, and there's no honest comparison with alternatives like Qdrant or Pinecone for any given use case. Notebook quality is uneven; the older entries (pre-2023) use deprecated client APIs and will throw errors without modification. There's no test suite or CI validation that notebooks actually run — the only signal something is broken is a GitHub issue. The directory structure has grown organically and is now a mess: `integrations/llm-agent-frameworks/data/` contains hundreds of Weaviate blog post MDX files that have nothing to do with the framework integrations they're nested under.