TechPrufer
Get started
Weaviate
Weaviate logoWeaviate · Feature trail

Weaviate Release Notes

3 posts · Mar 2026 – Aug 2026

What this trail covers

Weaviate v1.39 is released, making the Boost API and Maximal Marginal Relevance (MMR) diversity selection generally available. New previews include 4-bit Rotational Quantization and an experimental Search REST API. gRPC-Web support and HNSW snapshot rework are also included, improving disk usage and startup times. Performance improvements and community contributions are noted.

All Weaviate trails
  1. Weaviate 1.39 Release | WeaviateThis post details the general availability of the Boost API, which allows query-time rescoring based on filters, property values, and time/numeric decay. It also announces the general availability of MMR diversity selection, which balances query relevance with result diversity, and introduces 4-bit Rotational Quantization as a preview feature for vector compression. Additionally, it mentions the experimental Search REST API, gRPC-Web support, and HNSW snapshot rework for improved disk usage and startup times.
  2. Weaviate 1.38 Release | WeaviateIntroduces HFresh disk-based vector index (GA), MCP Server (GA), cluster-wide async replication (default), Boost API (Preview), and Nested Object Filtering (Preview). Also details performance improvements like production-ready replica movement, default vector index type setting, usage guardrails, and fractional BM25 property boosts.
  3. Weaviate 1.36 Release | WeaviateIntroduces HFresh, a new disk-based vector index in technical preview, inspired by the SPFresh algorithm. HFresh divides vectors into postings stored on disk, using an in-memory HNSW index over centroids for efficient search. It features incremental rebalancing, uses HNSW for centroid indexing, and employs Rotational Quantization (RQ) for compression. Key tunable parameters include searchProbe, replicas, and maxPostingSizeKB. It supports cosine and l2-squared distance metrics and is suitable for large datasets, cost-sensitive deployments, and write-heavy workloads. Server-side Batching is now generally available, using a persistent connection and dynamic backpressure for optimized throughput. Object TTL is also generally available, allowing expiration of objects based on creation time, last update time, or a date property, with configurable deletion schedules and an option to filter expired objects. Async replication is now generally available with collection-level configuration via the new `asyncConfig` parameter, offering fine-grained control over data comparison and propagation between nodes.