Exa

Exa just passed $100m in annualized revenue. We've grown almost exactly 10x per year for 3 years. We now handle 2 billion requests a month from 200k+ teams. And it’s still so early: We predict agents will search 1000x more than humans by 2030. Agents don't search like humans. They write long, complex queries and want all the world's data, not just the web. We've spent 5 years building a search engine designed for them. We now serve personal agents, AI labs, coding agents, GTM agents, really any agent. Exa gives these agents access to 100s of billions of web documents and tons of proprietary data. Our customers are increasingly noticing that information is the bottleneck to agent performance. While the world seems close to perfect intelligence with models like Fable and Astra, we are very far from perfect information. To our small but legendary team, thank you for getting us here, from engineering to research to GTM to ops. To our customers, it is an honor to serve you. We have some exciting releases coming soon and can't wait to share them 🫡

Exa is now one of the largest indexes in the world. We serve 80B pages, track 1.4T urls, and are on track to be Google-scale in early 2027. It's difficult to know the size of indexes, but we estimate Google to be ~1T, Bing to be ~500B, and Yandex ~200B. Brave mentioned in April that they're at 40B. Most of the web is trash that can hurt AI outputs, so you actually have to crawl way more than you serve and train models to filter out the trash. In terms of peak QPS, we estimate Bing at roughly 30k/s and Google at ~500k/s. In our case, because agent traffic often requires large fan-out (deep searches can use dozens of sub-searches) and can be spiky (like when AI labs RL with us), we're already starting to provision for Bing-scale traffic. Within 2 years, agents will search at many millions per second. They'll also want comprehensiveness over all data, far bigger than any index today. So the retrieval infra required will be larger than Google-scale in both dimensions.

We raised $85M. It's taken 4 years to get to this moment. Exa began as a gnawing itch. What if Google was outdated? What if you could use transformers to build perfect search over all the world’s information? What if we were the ones who did it? In 2021, this sounded crazy to most people I talked to. So we did YC, bought a million dollar GPU cluster, trained transformer models for years, made some serious breakthroughs, built a search API, grew our customer base to thousands, grew our team to 35 people, shipped the highest quality search for LLMs in the world (Exa API), shipped the most comprehensive search in the world (Websets), shipped the fastest search API in the world (Exa Fast), and now raised a Series B. Exa doesn’t sound so crazy to people anymore. With this funding, that original itch has turned into a feeling of duty. It’s not just a dream anymore. Exa is well-positioned to completely transform the way we find information as a society. The stakes could not be higher. As our species plunges headfirst into an increasingly chaotic AI world, misinformation will be rampant and quality information will become ever harder to find. Social media, news orgs, and traditional search engines are not equipped to handle the information overload from this new era. Exa is. Exa was built to be search infrastructure for this AI world. Unlike incumbents, we have no ads, meaning our incentives are purely to serve our users with the highest quality information. And unlike other startups, we spent years building our own search engine from scratch and making it customizable, meaning our users can get customized control over the world’s information. The next stage of Exa will be achieving the dream of perfect search. What this means is that we will handle literally any information request. Ex. “What’s the total number of articles about web search that came out in the past year”. To truly organize the world’s information, we’ll need to: 1. Gather all the world’s information 2. Develop methods to preprocess it such that it can be retrievable with any query no matter how complex The first step is building 10x larger scale indexing/processing infrastructure, training/researching novel embedding models on a 5x bigger GPU cluster, and customizing our product for tens of thousands of new customers around the world. Does this all sound crazy? Exactly. It’s the crazy ones who'll do it.

cooking our nextgen model on the Exacluster

Microsoft is shutting down the Bing API. This means that now, Google and Bing both don't have search APIs. Don't worry, there's a 5x cheaper option that's actually more powerful 😏