Turbovec: Google's TurboQuant for vector search in Rust
Turbovec – Google's TurboQuant for vector search in Rust
Turbovec is a Rust vector index with Python bindings, built on Google Research's TurboQuant algorithm. It compresses a 10M document corpus from 31 GB (float32) to 4 GB, while searching faster than FAISS. Key features include online ingestion without training, SIMD-optimized search (beating FAISS by 3.4× at 4-bit on average), incremental saves, and filter-at-search-time. It offers drop-in integrations for LangChain, LlamaIndex, Haystack, and Agno, and is designed for privacy-focused, air-gapped RAG stacks.
A 10 million document corpus takes 31 GB of RAM as float32. turbovec fits it in 4 GB - and searches it faster than FAISS.
- Eridrus
FAISS is no longer close to SoTA:
https://ann-benchmarks.com/index.html
- lmeyerov
Interestingly, while we don't fine-tune generative models for Louie.ai, we found fine-tuning embedding models to be a major $ saver. Instead of 1K-2K wide frontier embedding vector lens... Just 64. Huge savings on vector DB $$$.
I'm curious how that works with something like turboquant. Not needed any more, still dominant, better together, ... .
- ghm2199
Wow! 4GB for 10 million documents. This means one could build a reverse index much faster than before and devx processes like debugging, performance testing would become much smoother. Can't wait for the sqlite bindings to come out!
- nharada
It would be nice to have the README be a little more human written for a project where you actually want people to adopt it
- bobmarleybiceps
people should read turboquant's open review comments: https://openreview.net/forum?id=tO3ASKZlok