Sentence Transformers 6.0 shipped 18 Aug 2026: multi-vector index grows 7.5 MB to 311.5 MB, transformers floor moves to v5

Engineering · Aug 25, 2026 · 14 min read

Sentence Transformers 6.0 shipped ColBERT retrieval on 18 August 2026 and a 42x index

Sentence Transformers 6.0 published 18 August 2026 adds ColBERT-style late interaction. The gain is about one NDCG point on NanoBEIR. The cost is 311.5 MB where a dense index used 7.5 MB.

Manu Shukla

#embeddings #vector-search #RAG

Glowing 3D network of vector nodes rising from a server GPU, illustrating RAG embedding models

AI & Cloud Infrastructure · Jul 18, 2026 · 13 min read

4 RAG embedding models compared in 2026: Nemotron 3 Embed vs OpenAI, Cohere and open-weight

A senior-engineer comparison of RAG embedding models in 2026: NVIDIA Nemotron 3 Embed, OpenAI text-embedding-3, Cohere Embed v4 and open-weight Qwen3, with RTEB benchmarks, pricing and self-host break-even math.

Manu Shukla

#RAG #embeddings #Nemotron 3 Embed