The hidden cost of multi-vector retrieval: a ~16x larger index

tomaarsen · x · 2026-08-18

Multi-vector retrieval's performance edge comes with a real cost: 4,874 Natural Questions passages become 608,414 token vectors — 311.5 MB versus 20 MB for a simple 1024d dense index, about 16x. Three ways out: token pooling, a real late-interaction index, or using it as a reranker.

Related event: Sentence Transformers v6.0 ships with first-class late interaction models(33 posts)→

Original post →

More from coding & agent

coding & agent channel →