seekdb-js Changelog

• Release date: January 30, 2026
• Version: V1.1.0

Highlights

• More flexible collection names: Maximum name length increased from 64 to 512 characters to support longer, more descriptive names.
• Full-text search: New FulltextAnalyzerConfig for custom tokenizers and field mappings.
• Broader embedding support: Integration with 11 embedding platforms, including open-source and commercial models.
• Persistent metadata: Embedding function definitions can be stored in system tables for easier setup and reuse.
• table-fork (experimental): Snapshot-style copy at the collection level for experiments and rollback.

New features

Longer collection names

• Limit increased: Collection names can be 64–512 characters, so you can follow longer naming conventions.
• Compatibility: Existing collections with shorter names are unchanged.

Full-text index configuration (FulltextAnalyzerConfig)

• New configuration: The createCollection and getOrCreateCollection APIs accept a configuration argument. In addition to HNSW index settings, you can now specify full-text index configuration via FulltextAnalyzerConfig.
• Docs: Full-text index configuration.

Embedding function support

• 11 embedding platforms are supported, including:
  • Open source: sentence-transformers.
  • Commercial: OpenAI, Azure, AWS Bedrock, and more.
  • Full list: Embedding functions.

Table-Fork (experimental)

tip

In seekdb V1.1.0, FORK TABLE (copy collection) is provided as an experimental feature. It is not recommended for production use until it is declared stable in a later release.

• Fast collection snapshots: Create a copy of a collection in under a second.
• Typical use cases:
  • A/B testing with different index parameters
  • Data rollback and disaster recovery
  • See Fork a table for more scenarios.
• Performance: Fork uses copy-on-write (COW) at the storage layer. It completes in less than 1 second and does not depend on data size.