Mintplex-Labs/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2026-07-21 01:25:28 -04:00
Code Issues 328 Packages Projects Releases 10 Wiki Activity

[GH-ISSUE #4338] [FEAT]: RAG with Hybrid Search #2763

New Issue
Open
opened 2026-02-22 18:31:08 -05:00 by yindo · 8 comments
yindo commented 2026-02-22 18:31:08 -05:00
Owner
Copy Link
Copy Source

Originally created by @sha-roze on GitHub (Aug 29, 2025).
Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4338

Originally assigned to: @timothycarambat on GitHub.

What would you like to see?

I’d like to see support for RAG using hybrid search that combines both keyword-based and semantic vector search. This would allow AnythingLLM to retrieve more relevant context from documents by leveraging both traditional search techniques and modern embedding-based similarity.

Originally created by @sha-roze on GitHub (Aug 29, 2025). Original GitHub issue: https://github.com/Mintplex-Labs/anything-llm/issues/4338 Originally assigned to: @timothycarambat on GitHub. ### What would you like to see? I’d like to see support for RAG using hybrid search that combines both keyword-based and semantic vector search. This would allow AnythingLLM to retrieve more relevant context from documents by leveraging both traditional search techniques and modern embedding-based similarity.
yindo added the enhancementfeature request labels 2026-02-22 18:31:08 -05:00
yindo commented 2026-02-22 18:31:08 -05:00
Author
Owner
Copy Link
Copy Source

@sorincadar commented on GitHub (Sep 5, 2025):

I can't believe this is not a feature. I had to discover it in practice. I kept reading that the main power of Anything LLM is its RAG pipeline, but without this it is pretty weak tbh.

@sorincadar commented on GitHub (Sep 5, 2025): I can't believe this is not a feature. I had to discover it in practice. I kept reading that the main power of Anything LLM is its RAG pipeline, but without this it is pretty weak tbh.
yindo commented 2026-02-22 18:31:08 -05:00
Author
Owner
Copy Link
Copy Source

@xack20 commented on GitHub (Sep 26, 2025):

Please add this important feature.

@xack20 commented on GitHub (Sep 26, 2025): Please add this important feature.
yindo commented 2026-02-22 18:31:08 -05:00
Author
Owner
Copy Link
Copy Source

@Alminc91 commented on GitHub (Oct 2, 2025):

That would be amazing. LanceDB supports that feature. Maybe even with a rate we can set, like 75% semantic, 25% keyword weighing or so, with a simple algorithm or reranking (for that external local gpu models could be referenced). Exactly that is available in the new OpenWeb UI version, if it helps as reference.

@Alminc91 commented on GitHub (Oct 2, 2025): That would be amazing. LanceDB supports that feature. Maybe even with a rate we can set, like 75% semantic, 25% keyword weighing or so, with a simple algorithm or reranking (for that external local gpu models could be referenced). Exactly that is available in the new OpenWeb UI version, if it helps as reference.
yindo commented 2026-02-22 18:31:08 -05:00
Author
Owner
Copy Link
Copy Source

@Saswatsusmoy commented on GitHub (Oct 9, 2025):

@timothycarambat can I take up this issue?

@Saswatsusmoy commented on GitHub (Oct 9, 2025): @timothycarambat can I take up this issue?
yindo changed title from [FEAT]: RAG with Hybrid Search to [GH-ISSUE #4338] [FEAT]: RAG with Hybrid Search 2026-06-05 14:48:24 -04:00
yindo commented 2026-06-05 14:48:25 -04:00
Author
Owner
Copy Link
Copy Source

@SylvainVerdy commented on GitHub (Mar 5, 2026):

Hello,

I would like to know if you have made any progress on this?
best regards,
Sylvain

@Saswatsusmoy @timothycarambat

<!-- gh-comment-id:4005100262 --> @SylvainVerdy commented on GitHub (Mar 5, 2026): Hello, I would like to know if you have made any progress on this? best regards, Sylvain @Saswatsusmoy @timothycarambat
yindo commented 2026-06-05 14:48:25 -04:00
Author
Owner
Copy Link
Copy Source

@uncieray commented on GitHub (May 5, 2026):

I agree. I have tried using AnythingLLM to identify relevant content in a corpus of local documents with only limited success. RAG only doesn't work for common queries I want to do like this one:

"For the list of topics listed on page 4 in document XXX identify all documents and the sections within them that relate to these points. State how each section is relevant to each topic"

As a fallback I have had to pay to use Claude Cowork (which does it very well). My understanding is that hybrid search would provide this functionality. For what its worth my local setup is this: I have converted my PDFs to markdown, embedded them using mxbai-embed-large and now using Gemma 4 E4B for the query.

<!-- gh-comment-id:4377005430 --> @uncieray commented on GitHub (May 5, 2026): I agree. I have tried using AnythingLLM to identify relevant content in a corpus of local documents with only limited success. RAG only doesn't work for common queries I want to do like this one: _"For the list of topics listed on page 4 in document XXX identify all documents and the sections within them that relate to these points. State how each section is relevant to each topic"_ As a fallback I have had to pay to use Claude Cowork (which does it very well). My understanding is that hybrid search would provide this functionality. For what its worth my local setup is this: I have converted my PDFs to markdown, embedded them using mxbai-embed-large and now using Gemma 4 E4B for the query.
yindo commented 2026-06-05 14:48:25 -04:00
Author
Owner
Copy Link
Copy Source

@TeddyAfro commented on GitHub (May 14, 2026):

There are any Updates regarding this topic?
A hybrid search would significantly improve quality

<!-- gh-comment-id:4449464122 --> @TeddyAfro commented on GitHub (May 14, 2026): There are any Updates regarding this topic? A hybrid search would significantly improve quality
yindo commented 2026-06-05 14:48:26 -04:00
Author
Owner
Copy Link
Copy Source

@TeddyAfro commented on GitHub (May 27, 2026):

@Saswatsusmoy
thanks for the pull request. hope this will go soon into main

<!-- gh-comment-id:4555593258 --> @TeddyAfro commented on GitHub (May 27, 2026): @Saswatsusmoy thanks for the pull request. hope this will go soon into main
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
Desktop
Docker
Integration Request
Integration Request
OS: Linux
OS: Mobile
OS: Windows
UI/UX
blocked
bug
bug
core-team-only
documentation
duplicate
embed-widget
enhancement
feature request
github_actions
good first issue
investigating
needs info / can't replicate
possible bug
pull-request
Mirrored from GitHub Pull Request
 question
stage: specifications
wontfix
No Label enhancement feature request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: Mintplex-Labs/anything-llm#2763
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?