Exa Raises $250 Million Series C at $2.2 Billion Valuation to Build Search Infrastructure for AI Agents
Andreessen Horowitz leads a $250M Series C in Exa, valuing the AI-native search startup at $2.2B as it serves over 5,000 companies including Cursor, Cognition, and HubSpot.
Overview
Exa, the San Francisco-based startup building a search engine designed for artificial intelligence rather than human users, raised $250 million in a Series C round led by Andreessen Horowitz, the company announced on May 20, 2026. The funding values Exa at $2.2 billion, according to SiliconAngle, and comes as the broader market for AI-native infrastructure draws intensifying venture capital interest.
What We Know
Exa launched its AI-focused API in early 2023 and has since grown its customer base to more than 5,000 companies, according to PYMNTS. Notable clients include Cursor, Cognition, HubSpot, OpenRouter, and Monday.com. Over 400,000 developers have adopted Exa’s services, SiliconAngle reported. The company says its code search capabilities have recently improved, and the product is now used by “nearly every coding agent,” as PYMNTS noted.
Unlike conventional search providers, Exa operates its own independent infrastructure. CEO Will Bryk has argued that “Most other search providers actually wrap other search engines and therefore cannot compete on quality/latency/cost,” according to SiliconAngle. The company has built a custom vector database storing web embeddings queryable in one-tenth of a second, a proprietary data-processing platform called exa-d that parallelizes work across graphics cards, and custom embedding models trained on an in-house Nvidia cluster, per SiliconAngle. Its Exa Instant product completes queries in under 180 milliseconds.
Exa previously raised $85 million in a Series B round that included Nvidia Corp. and Y Combinator among its investors, according to SiliconAngle.
The new capital will be used to train next-generation search models and to scale infrastructure to support hundreds of thousands of searches per second, the company said, as reported by PYMNTS. Hiring in go-to-market functions is also planned, per SiliconAngle.
Market Context
Bryk has framed the company’s opportunity in expansive terms. “As trillions of agents come online over the coming years, search needs will grow thousands of times beyond the total search volume of Google,” he said, according to PYMNTS. He has also stated that “agents will need perfect search over all the world’s information at an unprecedented scale.”
The announcement arrived one day after Google unveiled what PYMNTS described as its most significant Search update in 25 years, which shifted the product toward AI-synthesized answers rather than ranked link lists. The timing highlights the competitive dynamic between established search incumbents and infrastructure startups positioning themselves as the retrieval layer for autonomous AI systems.
The Exa raise reflects a broader pattern in the venture market. As TechStartups noted, “the cost of high-precision, AI-native retrieval has become a primary driver of venture investment” in mid-2026, as investors concentrate capital on companies reshaping information access for the agentic era.
What We Don’t Know
Exa has not disclosed details about the specific technical benchmarks or independent evaluations that substantiate its performance claims relative to other search APIs. The company has not announced a timeline for reaching profitability, nor provided specifics on current revenue or query volumes. It is also unclear how quickly its independent infrastructure can scale to match the reach and freshness of indexes maintained by hyperscalers.