Micro1 raises funds at $500M valuation, eyes Scale AI’s market share
Micro1, a fast-growing data labeling startup, is raising funds at a $500 million valuation, backed by 01A and LG Tech Ventures, as it challenges Scale AI’s dominance amid a shift in the AI training ecosystem.
Micro1, a rising competitor to Scale AI, is closing a Series A fundraising round that values the startup at approximately $500 million, according to reports by Reuters.
Micro1 delivers AI-focused data-labeling services tailored for labs developing advanced machine learning systems. At the core of its value proposition is an AI-powered recruitment engine designed to onboard specialized experts, moving away from traditional crowdsourced labor models.
The company has reportedly seen its annualized revenue surge from $10 million early this year to over $50 million, and it expects to surpass $100 million by the end of September 2025.
Investors and board moves
The Series A round includes participation from investors 01A and LG Technology Ventures, according to the reports.
In a notable move, former Twitter COO Adam Bain has joined Micro1’s board, signaling the startup’s ambition to scale quickly and strategically.
Micro1’s momentum comes at a pivotal time for the data-labeling sector.
Scale AI, once considered the market leader, has faced turbulence after its CEO Alexandr Wang was recruited by Meta Platforms to lead its newly formed Superintelligence Labs.
That shift prompted high-profile clients, including Google and OpenAI to reconsider their partnerships with Scale AI amid concerns around research confidentiality in relation to Meta.
Broader landscape: Surge AI also on the rise
Micro1 isn’t the only challenger in this space. Another Scale AI rival, Surge AI, is reportedly raising up to $1 billion and reached over $1 billion in annual revenue last year.
AI labs are increasingly seeking bespoke, expert-led labeling solutions rather than generic task platforms. By focusing on quality and expertise, Micro1 is aiming to serve clients who require precision and domain-specific knowledge rather than volume alone.


