How to build Hybrid Search using Qdrant?

[karndeepsingh]

Hi,
I am trying to build hybrid search and I just caught up into Qdrant. I would like to understand following things:

1. Is qdrant free to use?
2. How I can use qdrant for building Hybrid Search? Any documentation or example would be better to get an idea on usage.
3. If I want to deploy this application on GCP using qDrant for Indexing , is it possible to do ??

Thanks

[generall]

Is qdrant free to use?

yes, it's Apache 2.0 license

How I can use qdrant for building Hybrid Search? Any documentation or example would be better to get an idea on usage.

Check out our latest article https://qdrant.tech/articles/hybrid-search/

If I want to deploy this application on GCP using qDrant for Indexing , is it possible to do ??

Yes, you can serve qdrant along your app in GCP, or access qdrant cloud from GCP

[karndeepsingh]

Thanks fore reply

Is qdrant free to use?

yes, it's Apache 2.0 license

How I can use qdrant for building Hybrid Search? Any documentation or example would be better to get an idea on usage.

Check out our latest article https://qdrant.tech/articles/hybrid-search/

If I want to deploy this application on GCP using qDrant for Indexing , is it possible to do ??

Yes, you can serve qdrant along your app in GCP, or access qdrant cloud from GCP

Thanks for replying.
is there any examples how I can implement Hybrid search with my dense model and sparse model using Qdrant?
It would be a great help if you can share examples of it or any reference that can guide on how to build a hybrid search using qdrant? I guess in article implementation examples are not mentioned.

[generall]

We don't have a sparse vectors serving yet, but it is a priority in our roadmap - https://github.com/qdrant/qdrant/blob/master/docs/roadmap/README.md

[alabrashJr]

@karndeepsingh I have not found any Hybrid Search related implementation yet, have you?

[nigh8w0lf]

Qdrant does not have Hybrid search (Keyword + Vector) feature for some reason, which is very disappointing ,you will have to Implement it yourself.
This forces me to consider Milvus as an option.