The problem
AI products need fresh, accurate web results, and search engines built for humans serve them poorly
What they do
Web search API built for AI agents and apps, on its own crawl index
Traction 5,000+ companies and 400,000+ developers; ARR ~$10M, up ~11x YoY (Sacra est., Sep 2025)
Hiring trend
- Open roles
- 57
- Ops roles
- 4
- Headcount
- ~150
~80 staff (Mar 2026) to ~150 (Sep 2026, per Exa job post); 56 open roles
Hiring in
- San Francisco
- New York
- London
Latest news
May 2026
Raised $250M Series C at $2.2B valuation led by a16z; LaunchDarkly president joins as CRO ↗Funding
- Last round
- $250M
- Stage
- Series C
- Date
- May 2026
Led by Andreessen Horowitz
- Hiring+257 open roles on the live job board (30+ roles = +2)
- Funding recency+2Series C · $250M, May 2026 (5 months ago): 6 months or less = +1.5, $100M+ round +0.5
- Traction+1.5Strong growth: 5,000+ companies and 400,000+ developers; ARR ~$10M, up ~11x YoY (Sacra est., Sep 2025)
- Red flags0None found in the last 24 months
- Total5.52 + 2 + 1.5 − 0 = 5.5 → Sprint (4.75 or more)
Distance (Half) is a separate read on how long the lead can last. It doesn't change the score.
Distance
Half · edge: learns from use
Its own 500B+ URL index, embedding models and GPU cluster are hard to copy versus search wrappers, but foundation labs bundling native search, plus Parallel and Tavily, make switching easy for many queries
Sources: Exa blog, May 2026: own crawl of 500B+ URLs, own models and GPU cluster ↗ · Sacra ex-employee interview, Feb 2026: about half of a customer's queries could move to SERP APIs or Parallel ↗ · Dealroom, May 2026: Google and incumbents are the main threat ↗ · Exa blog, Oct 2026: new Places index (211M places) beside Search, Contents, Agent and Websets ↗
Last checked Oct 25, 2026