# Integration of weaviate and langchain, how to use hybrid in v4 like as\_retriever in v3

**URL:** <https://forum.weaviate.io/t/integration-of-weaviate-and-langchain-how-to-use-hybrid-in-v4-like-as-retriever-in-v3/20030>\
**Category:** Support\
**Tags:** python\
**Created:** [March 18, 2025, 8:43pm UTC](https://forum.weaviate.io/t/integration-of-weaviate-and-langchain-how-to-use-hybrid-in-v4-like-as-retriever-in-v3/20030 "2025-03-18T20:43:59Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![cyf121232145](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.weaviate.io/cyf121232145/32/6917_2.png) [@cyf121232145](https://forum.weaviate.io/u/cyf121232145)\
**Post date:** [March 18, 2025, 8:43pm UTC](https://forum.weaviate.io/t/integration-of-weaviate-and-langchain-how-to-use-hybrid-in-v4-like-as-retriever-in-v3/20030/1 "2025-03-18T20:43:59Z")

</div>

### Description

I noticed that there is a method called as\_retriever in the v3 version, which is integrated into langgraph as a tool, but I can’t find anything similar in v4. The WeaviateVectorStore library that can be integrated as a retriever only supports vector retrieval.

### Server Setup Information

- Weaviate Server Version:  
Name: weaviate-client  
Version: 4.11.1

Name: langchain-weaviate  
Version: 0.0.4

---

<div class="post-metadata">

**Author:** ![DudaNogueira](https://yyz1.discourse-cdn.com/flex027/user_avatar/forum.weaviate.io/dudanogueira/32/7846_2.png) [@DudaNogueira](https://forum.weaviate.io/u/DudaNogueira)\
**Post date:** [March 19, 2025, 3:00pm UTC](https://forum.weaviate.io/t/integration-of-weaviate-and-langchain-how-to-use-hybrid-in-v4-like-as-retriever-in-v3/20030/2 "2025-03-19T15:00:01Z")

</div>

Hi @cyf121232145 !!

Welcome to our community 🤗

In the [new integration](https://github.com/langchain-ai/langchain-weaviate/), the similarity search already uses hybrid:

> <https://github.com/langchain-ai/langchain-weaviate/blob/56c874566c661a2e9c95b7332e882be2b66f4f3e/libs/weaviate/langchain_weaviate/vectorstores.py#L274>

Not that, because it passes all kwargs to `collection.query.hybrid`, you can use any of the parameters as documented here:

> **[Hybrid search | Weaviate](https://weaviate.io/developers/weaviate/search/hybrid)**
>
> Hybrid search combines the results of a vector search and a keyword (BM25F) search by fusing the two result sets.

Is this what you are referring to? Sorry, I don’t have much experience with langgraph.

Let me know if this helps!