- Added new search parameters and enhancements (e.g., prioritize matching fields, configurable hybrid weight, drop‑tokens mode, vector‑only on keyword results, group‑missing values, remote embedding timeout/tries, range faceting float supp...
- Fixed numerous bugs including HNSW index precision, analytics aggregation regression, embedding field prefix conflicts, reindexing on schema updates, group‑by with facet ranges, optional nested field handling, PaLM API downtime, infix de...
- Provided upgrade guidance due to data directory incompatibility for multi‑node clusters and downgrade constraints, with updated documentation link.