> For the complete documentation index, see [llms.txt](https://docs.flowiseai.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.flowiseai.com/integrations/langchain/retrievers/extract-metadata-retriever.md).

# Extract Metadata Retriever

This retriever is designed to automatically extract keywords from query. The extracted JSON output is used as metadata filter for vector store.

For example, when we ask a question: "What is the profit for Apple", LLM will give an output of `{source: "apple"}`, and this will be passed to vectore store's metadata filter.

<figure><img src="https://823733684-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F00tYLwhz5RyR7fJEhrWy%2Fuploads%2F78aZPFLaSnpTWa2SQEWO%2Fimage.png?alt=media&amp;token=8189a1b3-cac2-4e93-b883-98f853dd3745" alt=""><figcaption></figcaption></figure>