v9.6.0

neuml/txtaiv9.6.0Feb 25, 2026by davidmezzetti

AI Summary

This release adds many great new Embeddings database features including BB25 normalization, ngram tokenization, database expression indexes, and conditional JSON data storage.

Key Highlights

Bayesian normalization 'bb25' method for keyword scoring
BB25 normalization for sparse encoders
Log-odds conjunction fusion for BB25 hybrid search
Keyword ngram tokenizer
Database expression indexes and store field to columns configuration

New Features

Add Bayesian normalization 'bb25' method for keyword scoring
Add BB25 normalization for sparse encoders
Add log-odds conjunction fusion for BB25 hybrid search
Add keyword ngram tokenizer
Add support for custom SQL functions with DuckDB
Add deterministic setting to database functions
Add ability to index database expressions
Add store field to columns configuration

Full Release Notes

**This release adds many great new Embeddings database features: BB25 normalization, ngram tokenization, database expression indexes and conditional JSON data storage**

See below for full details on the new features, improvements and bug fixes.

New Features
--------------------------
- Add Bayesian normalization "bb25" method for keyword scoring (#1023, #1037, Thank you @sigridjineth)
- Add BB25 normalization for sparse encoders (#1042, #1046 Thank you @jaepil)
- Add log-odds conjunction fusion for BB25 hybrid search (#1041 Thank you @jaepil)
- Add keyword ngram tokenizer (#1017, #1038)
- Add support for custom SQL functions with DuckDB (#1047)
- Add deterministic setting to database functions (#1048)
- Add ability to index database expressions (#1049)
- Add store field to columns configuration (#1050)

Improvements
--------------------------
- Cache Agent Template (#1051 Thank you @HemantSudarshan)

Bug Fixes
--------------------------
- Fix: escape single quotes in dynamic column names (#1026 Thank you @RinZ27)
- Fix: cursor error with multithreaded keyword search (#1039)