Qdrant's Constella lets you swap query embedding models without re-embedding docs
qdrant_engine · x · 2026-09-29
Qdrant released Constella, a research preview that lets you change your query embedding model without re-embedding all documents. Its Nano and Zero models are trained to produce vectors compatible with Stella's document embeddings: Zero is extremely lightweight, while Nano is a small transformer approaching Stella's retrieval quality—letting teams trade off latency vs. quality on the fly.
Related event: Qdrant Launches Constella Preview: Swap Query Models Without Re-embedding(3 posts)→
More from Infra
- Swapping matmul for associative-algebra layers boosts 110M LM throughput 7.8% — Ilya Koziev · 2026-09-29
- Shaw mocks data center opponents: hating compute while using the internet is incoherent — zealcaiden · 2026-09-29
- Modeling 1B agent VMs by 2030: what personal AI agents mean for CPU demand — AccBalanced · 2026-09-29
- Tessera: retrieval-driven KV cache reuse cuts RAG serving TTFT by up to 3.6x — _reachsumit · 2026-09-29
- Venice's tokenized inference-credit model is being copied — and oversupply looms — 0xJeff · 2026-09-29
- Model Casting: Mid-Training Recipe Sparsifies FFN Activations for Fewer FLOPs — francoisfleuret · 2026-09-29