AI Data Labeling Rising Star Micro1 Raises $35 Million, Valuation Soars to $500 Million

September 12United States of AmericaAI data labeling startup Micro1 announced the completion of a $35 million (roughly Rs. 250 million) Series Afinancing, with a post-investment valuation of $500 million (equivalent to about 3.5 billion yuan).The round was led by 01 Advisors, an organization co-founded by former X CEO Dick Costolo and former X COO Adam Bain.

According to TechCrunch, Micro1 has been around since 2022 and its main business is enterpriseData annotation and training manpower contracting management(math.) genusRelying on AI Interviewer Zara recruits experts in various fields at scale and combines them with its own data to deliver labeled and reviewed data.In June 2025, after Meta invested $14 billion (equivalent to about 99.7 billion yuan) in Scale AI and hired its CEO, Alexandr Wang, core players in the AI field such as OpenAI and Google announced the termination of their cooperation due to concerns about the possible leakage of data to Meta. However, these companies are still in dire need of data labeling support, so Micro1 is taking advantage of the situation to take on orders.

I. Annual recurring revenue amounted to $50 million, more than seven times the amount at the beginning of the year

According to the company's announcement, the $35 million (equivalent to about 250 million yuan) funding in this round will be used toExpansion of expert networks and development of new training environments.Bain, co-founder of investor 01 Advisors, will join the board of directors, with DoNotPay founder Joshua Browder also on board. Bain commented, "Micro1 has demonstrated unprecedented speed in delivering new artificial data to cutting-edge labs."

Micro1 CEO Ali Ansari, 24, revealed, "The company's annual recurring revenue (ARR) has now reached $50 million (Rs. 356 million), much higher than the $7 million (Rs. 50 million) at the beginning of the year."

▲Ali Ansari, 24, founder of Micro1

Of course, this is still a long way from the competition: For example, Mercor (also an AI data outsourcing platform that provides labeling/evaluation manpower and management to model developers) has an ARR of more than $450 million (roughly Rs. 3.2 billion); and Surge (a data service provider that focuses on high-quality expert labeling and complex task evaluation) is reported to be generating revenues of $1.2 billion in 2024 (roughly Rs. 8.5 billion). However, Micro1's market size growth and rate of product adoption in the AI data space looks to be steadily increasing.

Secondly, taking high-quality labeling for specific areas, the company disclosed three core strengths

Unlike earlier Scale AI, which relied on low-cost labor, Micro1 emphasizesDomain-specific high-quality labeling, i.e., experts such as software engineers, doctors, professional writers, etc. are involved in data production and validation. As a result, Micro1 has createdAI Recruiter Zara, which is used to screen expert contractors at scale, officially says it has covered thousands of top talent from Stanford, Harvard and other colleges and universities, and is continuing to expand.

▲zara product homepage

Micro1's official blog mentions that there are "three pillars" to its approach:

1)AI interviewer (AI interviewer): Human intelligence screening through machine-assisted interviewing, looking at more than just resumes and accurately matching candidates to more suitable positions;

2)Talent performance management (TPM): Sink real performance data on the job to feed predictive modeling and QC processes;

3)Data platform for training frontier AI models (Data platform for training frontier AI models): Already serves some of AI's core players and the Fortune 100 for high-value workflows such as model evaluation and training.

The AI Lab is reportedly seeking"Environmental" services, i.e., performing simulated tasks in a virtual workspace to train an AI agent.Currently, Micro1 has embarked on the development of a related product. It is worth noting that it is difficult for any one company to handle all the data needs of an AI lab, so labs commonly adopt a multi-vendor strategy to spread out dependencies.

Conclusion: The Data Labeling Industry is Shifting to Refinement

As can be seen from Micro1's business layout, the focus of competition in the labeling industry is now shifting from the pursuit of scale alone to a focus on quality and efficiency, and as a result, the cost structure of the services it provides and the shape of its products are also changing.

For AI data labeling enterprises, in the future, who can form a complete business system in the areas of expert supply, automated recruitment and reusable environment will be more likely to occupy a larger market in the new round of data service pattern.

Source: TechCrunch