BENCHMARKS

[logxdx]

Benchmark Helix-db against other vector databases in the market.

[Gand4lf]

Hey apologies for the delay,

We've benchmarked ourselves against qdrant, and for the same dataset of 1M vectors our vector search is ~5ms compared to their ~3ms. However, our vector implementation runs on disk, while theirs runs on memory.

So you get almost the same performance and use less compute/energy.