- Performance and reliability upgrades: deferred point updates, non‑blocking Gridstore flushes, faster filtered search, improved WAL error handling, and reduced lock contention during shard transfers.
- New observability and API features: request tracing IDs in audit logs, per‑collection Prometheus metrics, audit‑log read endpoint, configurable max batch size in strict mode, and Edge crate/Python packages for Qdrant Edge.
- Security and stability fixes: enforce snapshot recovery from snapshot directory, multiple WAL replay/persistence bug fixes, GPU Vulkan name pointer correction, and numerous panic and shard‑related bug resolutions.