- Added extensive search enhancements including union/merge of results, dictionary‑based stemming, random sorting, hybrid re‑ranking, pivot and decay function sorting, field‑level token separators, bucketing, collection truncation, geo‑pol...
- Introduced new API endpoints for alter‑schema status and improved faceting performance, alongside stability and performance fixes for joins and high‑concurrency scenarios.
- Fixed numerous bugs such as async reference restoration, nested reference sorting, conversation API edge cases, default sorting scores, vector query validation, embedding field type checks, race conditions, and hybrid search cutoff issues.