Home / Directory / Agents & orchestration / AI web search APIs / Exa
Exa
Search engine and API from San Francisco built for AI agents, with neural web search, page contents, deep research, and people search. Exa raised a $250 million Series C led by Andreessen Horowitz in May 2026 at a $2.2 billion valuation.
AI product teams building agents, coding assistants, and research tools that need high-quality web results and page text from a single API.
Teams that only need Google-style results for SEO tracking, or buyers who want one flat monthly price.
Verdict
Exa is a search company in San Francisco, started in 2021 by Will Bryk and Jeff Wang, who met as freshmen at Harvard. AI companies such as Cognition, CodeRabbit, and HubSpot use its API to search the web, pull clean page contents, run multi-step deep research, and find people and companies. It stands out because Exa crawls and indexes the web itself and trains its models to predict useful links, so an agent can ask in plain language and get pages that match the meaning rather than the keywords.
Score Breakdown
How Exa scores in the categories that matter to its buyers.
Pricing
Exa publishes usage prices per thousand requests. A free tier comes with monthly credits and no payment method.
| Plan | Price | What stands out |
|---|---|---|
| Starter | $0 | $10 of credits every month plus a $10 onboarding bonus |
| Search | From $4 / 1,000 requests | Web search tool calls for agents |
| Contents | $1 / 1,000 pages | Clean page text for results |
| Deep Search | $12–15 / 1,000 requests | Multi-step agent workflows |
| Enterprise | Custom | Volume pricing, custom data providers, enterprise controls |
The Field at a Glance
How Exa compares with other AI web search APIs vendors we reviewed, by Overall Score and relative typical engagement cost.
Exa scores 7.7 in AI web search APIs, under Firecrawl (7.9), and ahead of Linkup (7.5) and Valyu (7.2). Relative cost lands mid-band for this subcategory.
Use-case matrix
| Use case | Fit | Notes |
|---|---|---|
| Coding agents and assistants | Strong | Cognition says Exa powers all parts of Devin. |
| Research and deep search | Strong | Deep Search runs multi-step queries for agents. |
| People and company search | Strong | Exa indexes more than a billion people profiles. |
| Code review context | Strong | CodeRabbit reports one Exa search replacing several from its old provider. |
| SEO rank tracking | Weak | Exa does not mirror Google's result pages. |
| Fixed monthly budgets | Mixed | Usage pricing varies by endpoint and depth. |
Review Excerpts
Excerpts from company materials and customer stories for Exa, plus a representative buyer view drawn from public information. Paid reviews and pay-for-play sites were excluded.
“Traditional search engine solutions weren't enough for us. Exa powers all parts of Devin.”
“Exa consistently shows higher quality than other AI search providers we tested. What Exa is able to do in 1 search took our old provider multiple searches, taking 5 times as long.”
“We evaluated five different search providers, but none came close to Exa's quality and breadth of data.”
“Cost scales per thousand requests, and deep research and contents calls cost more than plain search, so a high-volume agent needs a budget.”
Representative buyer view
Where customer reviews couldn't be found, our analysts attempted to give a voice to buyer sentiment, drawing on real information from review sites, forums, and product documentation.
Based on: Exa’s pricing page
Company Facts
- Founded
- 2021 (Y Combinator)
- Headquarters
- San Francisco (Y Combinator)
- Employees
- About 75 (Y Combinator)
- Accelerator
- Y Combinator, Summer 2021 batch (Y Combinator)
- Website
- exa.ai
Key Leadership
Jeffrey Wang
Co-founder
Will Bryk
Co-founder and CEO
Methodology
This page is an independent evaluation of Exa for buyers comparing options in AI web search APIs. AI Industry Reviews accepts no sponsorships, advertising, or pay-for-placement fees. Exa did not pay for this review.
What we scored
The headline number is an Overall Score on a 0-10 scale. Eight criteria sit under it in two groups.
- Buyer outcomes (for AI web search APIs)
- Result quality for agents
- Speed & one-call coverage
- Breadth of endpoints
- Cost predictability
- Company & commercial
- Innovation & product leadership
- Project management & communication
- Pricing
- Contract fairness
Pricing measures whether the price looks fair for the value delivered, including packaging and renewal friction that show up in real buying cycles.
Score Composition
| Input | Weight | What it covers |
|---|---|---|
| Reviews | 40% | A proprietary read of what practitioners say about likes, complaints, and day-to-day use, including public review sites, forums, and private chat rooms. Paid reviews and pay-for-play sites are out of scope. |
| Product | 35% | Hands-on look at screens and workflows. |
| Pricing | 15% | Whether the price looks fair for what you get. |
| Docs & training | 10% | Docs, tutorials, and training material. |
How we balanced the evidence
The Overall Score is the simple average of the eight criteria. Recommendation language follows that score and the use-case matrix above.
Scope
Exa is graded here as AI web search APIs. Criteria scores can move as more review volume and product checks are added.