Hybrid SQL Search | Spice AI

Hybrid SQL Search

Combine vector similarity, full-text, and relational filters in a single SQL engine. Run hybrid search across your data-no extra pipelines or systems needed.

Find what matters with hybrid search

Scale to billions of embeddings and maintain control with SQL filters, joins, and ranking in a single runtime.

Relevant results, every time

Retrieve answers that reflect context (semantic meaning) and precision (exact match).

Search at any scale

Index and search billions of embeddings with low-latency, across cloud object storage or operational databases. Performance remains consistent as datasets grow.

Full SQL control

Design, refine, and combine search results entirely in standard SQL.

Build for scale with tools developers trust

A familiar, SQL-powered experience enables advanced search without leaving existing SQL workflows. Spice provides full transparency, control, and scalability with modern infrastructure behind the scenes.

Compose hybrid search in SQL

Write, filter, join, and re-rank using standard SQL commands like vector_search and text_search. Integrate advanced search directly into your analytics or application stack.

Serve billions of records from cloud storage

Store, index, and retrieve embeddings with partitioned S3 vector indexing. Spice automatically manages ingest, refresh, and metadata for scalable, efficient queries.

Fine-tune results with built-in controls

Spice hybrid search is fully expressed in standard SQL, letting you filter, join, re-rank with reciprocal rank fusion, and project fields within each query with zero middleware.

Trusted in production

Teams use Spice to replace complex search stacks with one engine that's fast, scalable, and controllable-without standing up additional systems.

“Spice AI grounds AI in our actual data, using SQL queries across many data sources. This brings accuracy to probabilistic AI systems, which are very prone to hallucinations.”

Rachel Wong

CTO, Basis Set

“Partnering with Spice AI has transformed how NRC Health delivers AI-driven insights. By unifying siloed data across systems, we accelerated AI feature development, reducing time-to-market from months to weeks - and sometimes days. With predictable costs and faster innovation, Spice isn't just solving some of our data and AI challenges - it's helping us redefine personalized healthcare.”

Tim Ottersburg

VP of Technology, NRC Health

Integrations across all of your data sources

Accelerate your data stack with a library of 30+ prebuilt connectors for the most common databases, warehouses, and file stores-from Databricks and S3 to MySQL and PostgreSQL.

FAQs

How does hybrid search work?

Hybrid search in Spice.ai merges vector similarity (semantic) and full-text BM25 (keyword) results into one ranked output. Both search types run in parallel, and their ranks are combined using Reciprocal Rank Fusion (RRF) for optimal relevance.

How is Spice.ai different from a vector database?

Traditional vector databases require you to pair a vector index with separate text and keyword search systems, all running on clusters you have to provision and maintain. Spice.ai unifies vector, text, and relational search in a single runtime that you can deploy locally, in your cloud, or fully managed.

When should I use S3 Vectors?

Use S3 Vectors when you need to store and query embeddings at large scale. It's ideal for workloads with millions or billions of vectors that don't need always-on compute.