Skip to main content
usedby
Search tools, companies

How Exa uses Zilliz

<200ms

Neural search latency with Exa Instant, reduced

As published on zilliz.com. Captured by usedby on Oct 6, 2026.

What happened

Exa uses Zilliz Cloud (managed Milvus) to power its entity search layer for companies, people, and code, serving as the primary index and recency cache. It handles hybrid dense and sparse vector search with metadata filtering and frequent upserts, feeding Exa's Search API and Websets product.

Summary written by usedby from the source page, in English. The figures are those of Zilliz and Exa, not ours.

Hybrid search combining dense vectors, sparse vectors, RRF reranking, and metadata filters in a single API call. Exa Instant reduced neural search latency from seconds to under 200ms

From the page. zilliz.com, captured Oct 6, 2026

Zilliz Cloud has been an important part of Exa’s journey to build and scale entity search, giving us the retrieval performance and operational simplicity we need to scale quickly and confidently.

Jeffrey Wang, Co-Founder, Exa. Source, captured Oct 6, 2026

Zilliz gave us real-time retrieval for our AI search system at scale with tight latency targets. It freed up engineering cycles and let us focus on improving reasoning on the model side, not managing infrastructure.

Jeffrey Wang, Co-Founder, Exa. Source, captured Oct 6, 2026

We believe AI agents will become a fundamental interface for how people work, learn, and make decisions, and that only happens if those systems can access real-world information with speed, precision, and trust.

Jeffrey Wang, Co-founder of Exa, Exa. Source, captured Oct 6, 2026

What the story claims, and what we checked

We compared the story with its live page on Oct 6, 2026.

  • The figure: <200msCheckedThe figure is on the page; its label is our wording.
  • The passage quoted aboveCheckedCopied word for word from the page, near the name of Exa.
  • Exa uses ZillizCheckedConfirmed line. Latest check across sources: Oct 6, 2026.
  • The result itselfNot checkedWe quote it; we did not measure it.

Same company, same tool or same industry.

~45 msP99 dense retrieval latency in production acrossConsensus uses Zilliz. Another customer of Zilliz60-80%“60-80% of the time spent on consuming, digesting, and identifying relevant data points”Filevine uses Zilliz. Another customer of Zilliz< 10msSearch latency with high recallOpenEvidence uses Zilliz. Another customer of Zilliz