In the Press

Product updates

Blog

Serving BM25 with 50% memory reduction using a novel encoding scheme

Serving BM25 with 50% memory reduction using a novel encoding scheme

How we cut our BM25 index footprint in half at billions‑document scale without sacrificing performance.

How we're building the next generation of search

How we're building the next generation of search

Before exploring other worlds, we should fully understand our own

Search that’s >20x better than Google for complex queries

Search that’s >20x better than Google for complex queries

A big step toward perfect search

State of the art web search API

State of the art web search API

The best search engine designed for LLMs

A Perfect Search Engine

A Perfect Search Engine

Before exploring other worlds, we should fully understand our own

How we built a web-scale vector database

How we built a web-scale vector database

It uses clustering, matryoshka embeddings, binary quantization, and SIMD operations. Written in rust of course 🦀

Exa raises Series A to build the search engine for AI.

Exa raises Series A to build the search engine for AI.

Funding led by Lightspeed Venture Partners will allow Exa to scale their first search product and become the data layer for AI applications.

We need superknowledge before superintelligence

We need superknowledge before superintelligence

The most important technical problem of our​​ time

Scaling our highlights server

Scaling our highlights server

Serving real-time embeddings at scale is challenging. To launch Exa Highlights, we 4X’ed throughput by migrating from Python to Rust.

Announcing Exa

Announcing Exa

Same mission, new name

Building Search for the Post-ChatGPT World

Building Search for the Post-ChatGPT World

Our information ecosystem is broken, and the best way to fix it is to combine LLMs with high quality content from the Internet