Starstruk Logo MarkStarstruk

ts-v3.0.7

mem0ai/mem0ts-v3.0.7Jun 10, 2026by whysosaket

AI Summary

Node SDK update adding LM Studio embeddings and fixing score normalization.

Key Highlights

  • Added `LMStudioEmbedding` provider for local embeddings via LM Studio server.
  • Added opt-in `explain: true` option to `Memory.search()` for score breakdowns.
  • Parallelized entity boost searches to eliminate latency.
  • Fixed score inversion in the Redis vector store adapter.

New Features

  • LM Studio embedding provider for local embeddings.
  • Explain mode for score breakdowns (semantic, keyword, entity, temporal) in search results.
  • Parallel entity boost searches for faster performance.
  • Fixed OpenAI embedder encoding format to `float`.

Full Release Notes

## Mem0 Node SDK (v3.0.7)

**New Features:**
- **Embeddings:** Add `LMStudioEmbedding` provider for local embeddings via the LM Studio server ([#5377](https://github.com/mem0ai/mem0/pull/5377))
- **Memory:** Add opt-in `explain: true` option to `Memory.search()`. When enabled, each result includes a `scoreBreakdown` object with `semantic`, `keyword`, `entityBoost`, and `temporalBoost` fields so callers can inspect and tune retrieval ranking ([#5102](https://github.com/mem0ai/mem0/pull/5102))

**Bug Fixes:**
- **Memory:** Parallelize entity boost searches in `Memory.search()`. All entity embed + store lookups now run concurrently instead of sequentially, eliminating multi-second latency on entity-rich queries with remote embedding providers ([#5377](https://github.com/mem0ai/mem0/pull/5377))
- **Vector Stores:** Normalize similarity scores to `[0, 1]` (higher = better) — fixed score inversion in the Redis vector store adapter ([#5391](https://github.com/mem0ai/mem0/pull/5391))
- **Embeddings:** Request `encoding_format: "float"` from the OpenAI embedder in both `embed()` and `embedBatch()`. Fixes incorrect vector dimensions when using OpenAI-compatible proxies that default to base64 encoding ([#5170](https://github.com/mem0ai/mem0/pull/5170))