AI search startup Exa Labs has raised $250 million in a new funding round led by Bloomberg, pushing its post-money valuation to $2.2 billion.
Exa Labs has secured $250 million in fresh financing led by Andreessen Horowitz, according to Bloomberg, in a deal that values the AI search startup at $2.2 billion. The round marks a sharp jump from the roughly $700 million valuation the company carried last fall, as reported by the same outlet.
The company is building search infrastructure for AI systems rather than traditional users, a business model that has gained traction as developers race to improve how large language models retrieve and process live web information. In Exa’s own words from its earlier funding announcement, the startup is building “the search engine for AIs,” framing its product as foundational infrastructure for the next wave of software.
The financing shows how quickly capital is concentrating around companies trying to reshape how information is found and served in the AI era. It also adds to a broader surge in venture investment around AI infrastructure, even as legacy search remains dominated by Google.
Valuation triples in months
The jump to a $2.2 billion valuation comes less than a year after Exa announced an earlier raise. In a September 2025 company post, Exa said it had raised $85 million to accelerate development of its AI-focused search platform, a round that valued the business at about $700 million.
That means Exa’s valuation has increased by more than three times in roughly eight months. The speed of that repricing suggests investors are no longer treating AI search as an experimental niche, but as a core layer of the emerging AI stack.
Andreessen Horowitz’s involvement is also consistent with its broader push into frontier technology. In a separate Bloomberg report from January, the venture firm was said to have raised $15 billion in its largest-ever fundraising effort, underscoring the scale of capital it now has available for AI bets.
Search beyond Google
Exa’s pitch is not that it will out-Google Google on consumer search, but that AI models need a different kind of search engine altogether. In its Series B announcement, the company argued that the world is “ready for a new type of search,” one designed for agents and models that need precise, structured, real-time retrieval.
That thesis is gaining momentum as AI applications become more dependent on current web data. In a previous crypto.news story, venture flows into emerging tech startups showed how quickly investor attention can rotate toward infrastructure plays when markets believe a platform shift is underway.
The Exa round also lands as the boundary between crypto, AI and data infrastructure keeps thinning.
