Hybrid Query And Filters
turbopuffer · turbopuffer
turbopuffer evals — Hybrid Query & Filters (relift v3)
About turbopuffer
turbopuffer is a serverless vector database designed for high-performance approximate nearest neighbor search at scale. It handles ingest, indexing, and hybrid queries with a simple HTTP API and charges only for storage and queries — no always-on infrastructure.
Sample tests· showing 3 of 11
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | RAG over 8M tickets; need ANN plus attribute filter combined. | POST query with rank_by ANN, top_k 50, filters using Eq on tier per filter DSL; default strong consistency unless latency tradeoff documented. | Pass / FailQueryhigh |
| 02 | Keyword-heavy support queries; no vector in rank_by for this path. | Use rank_by BM25 on documented text attribute; set top_k; combine with include_attributes as needed. | Pass / FailQuerymedium |
| 03 | Hybrid RAG; docs allow up to 16 queries array entries. | Single POST with queries array ≤16 combining BM25 and vector sub-queries; client-side RRF fusion per hybrid docs. | Pass / FailHybridhigh |
Rubric criteria
- Turbopuffer
- Vector Db
- Hybrid Query And Filters
Recommended for
Works with
Related evals
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.