micro1, a data lab that trains frontier AI models and evaluates agents, said on 12 October 2026 that it has surpassed $1 billion in annual run rate while remaining profitable. It also said that after passing $500 million in run rate this summer, it priced its Series B at a $4 billion valuation, alongside an approximately $25 million tender for early employees.
The round was led by Dream Ventures, with Microsoft, Javelin, Flybridge, G Squared, Up Partners and LG, according to the company’s blog post and press release.
Business Pill · RUN-RATE REVENUE
A one-minute explainer of run-rate revenue: annualising a recent period’s revenue to estimate a yearly figure. It teaches the general idea only and says nothing about any organisation in this story.
The key insight: As we read it, expert training data is now a billion-dollar line of business: micro1 says it passed $1 billion in annual run rate with about 180 core staff, and that it priced a Series B at a $4 billion valuation after passing $500 million this summer.
What micro1 Announced
In the post, CEO Ali Ansari writes that “today, i’m proud to share we’ve surpassed $1B annual run rate while remaining historically profitable.” The release describes it as maintaining a record of profitability.
micro1 says the tender gives its early employees liquidity. Its post describes Dream Ventures as “a longtime supporter and one of our earliest investors.” Neither the post nor the release gives the size of the Series B round.
The post also names individual backers, including Guodong Zhang and Igor Babuschkin, both described as co-founders of xAI, DoorDash CTO Andy Fang and Jerry Tworek, described as an ex-VP of Research at OpenAI.

A Small Team and a Hard Year
micro1 says its core team is about 180 people across more than 14 countries and fully remote, and that it intends to slow team growth significantly from here.
The post says that about three years ago micro1 “lost 60% of its revenue overnight and almost went bankrupt.” It says it has kept close to 100% core team retention.
What micro1 Sells
On its website, micro1 lists Realm, “Frontier evaluations and RL environments for AI Labs”, Cortex, an improvement layer for agentic AI, and expert-demonstrated data for robotics labs.
In the post, micro1 lists eight principles, including working with “the tens of thousands of experts helping train models” and a commitment to “refuse to work with adversarial nations.” It says it will invest in data generation models to make datasets and RL environments less expensive.
On 9 October we covered micro1’s plan to spend $1B over 12 months licensing company data for reinforcement learning environments, financed with capital from Citi and Hercules Capital.
What micro1 Says Comes Next
Ansari writes that “data is on its way to becoming the largest industry humanity has ever seen.” micro1 says it aims to become the leading company in the data industry.
The Structural Read
As we read it, the growth sits in what labs buy after pretraining: micro1 lists frontier evaluations and RL environments for AI labs among its products, and says tens of thousands of experts help train models.
As we read it, the team size is part of the pitch: micro1 says it reached $1B run rate with about 180 core people and intends to slow team growth significantly.
As we read it, the 9 October data commitment and today’s update fit together: micro1 says it is investing in data generation models to make datasets and RL environments less expensive.
Ali Ansari, CEO, micro1, end of Q3 update, 12 October 2026
“data is on its way to becoming the largest industry humanity has ever seen, and we’re certain this is just the beginning.”
Three Implications
FRONTIER LABS micro1 lists frontier evaluations and RL environments for AI labs, and says it works with tens of thousands of experts helping train models.
INVESTORS micro1 says Dream Ventures led its Series B at a $4 billion valuation, with Microsoft, Javelin, Flybridge, G Squared, Up Partners and LG.
EARLY EMPLOYEES micro1 says an approximately $25 million tender gives its early employees liquidity.
The Business Engineer Lens
This story maps onto the Business Engineer framework The Four Intelligence Moats.
The framework puts it this way: “The Verifier Moat โ built by RLVR on domain-specific reward signals, building into a new asset class for reasoning-heavy verticals.”
As we read it, micro1’s numbers show the supply side of that moat: it sells RL environments and expert data to AI labs, and says that business has passed $1 billion in annual run rate.

What Is Not Established
Annual run rate is micro1’s own metric, not audited revenue, and the company does not say how it is calculated. The round size and the date the Series B was priced are not given, beyond “after surpassing $500M run rate this summer.” The profitability claim is the company’s.
The Bottom Line
micro1 says it has passed $1 billion in annual run rate with about 180 core staff, and that it priced a Series B at a $4 billion valuation led by Dream Ventures, with Microsoft and LG among the investors.
95,000+ executives read Business Engineer for the AI strategy frameworks cited by ChatGPT, Claude, and Perplexity.
A note on sourcing. On 12 October 2026 we read micro1’s blog post and its press release of the same day in full. Annual run rate, profitability and the valuation are the company’s own statements, not audited figures, and the round size is not given. Nothing here is investment advice.
Sources: micro1: end of Q3 update (blog post by CEO Ali Ansari), 12 Oct 2026 · micro1 press release via EIN Presswire, 12 Oct 2026









