typesense / typesense/typesense
Vector search: search for multiple embedding-vectors
Open
Nobody has claimed this yet.
feature:vectorsearch
- Dominant language
- C++
- Stars
- 26.6k
- Forks
- 973
- Avg merge
- 18h 45m
- Merged PRs (30d)
- 4
Description
Description
Is it possible to search for multiple embeddings at the same time? In my specific case, I want to make a similarity search for text and image embedding. I am aware of hybrid search, where multiple sources are being taken, but I am unaware of a multi-vector search.
What currently works (single):
{
collection: posts,
q: *,
vector_query: "image_embedding:([], id: XXX)",
page: 1,
per_page: 8,
}
What I would like to do (multi-vector):
{
collection: posts,
q: *,
vector_query: "image_embedding:([], id: XXX), text_embedding:([], id: XXX)",
page: 1,
per_page: 8,
}
Typesense Version:
v0.25.1
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reviewing the existing single-vector vector_query behavior described in the issue and the Typesense v0.25.1 vector-search implementation. Define how text and image embeddings should be accepted and combined in one query, then verify that multi-vector requests return the expected similarity results.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- search
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100