AI infrastructure startup Mercor has raised $350 million in Series C funding, quintupling its valuation to $10 billion, the company confirmed to TechCrunch.
The round was led by Felicis Ventures, with participation from Benchmark, General Catalyst, and new investor Robinhood Ventures.
The raise follows Mercor’s $100 million Series B in 2024 — also led by Felicis — which valued the company at $2 billion. This latest round cements Mercor as one of the fastest-growing AI infrastructure firms of 2025, with sources noting the company received multiple competing offers from investors.
“AI is advancing at incredible speed, but it still struggles with the subtleties of economically valuable work — understanding intent, taste, and trade-offs,” the company shared in a blog post announcing the raise. “Our mission is to close that gap by bringing human intelligence back into AI development.”
From AI Hiring Platform to Human–AI Training Network
Founded in 2022, Mercor began as an AI-driven hiring marketplace before pivoting to a new model: connecting AI labs with domain experts — including scientists, doctors, and lawyers — who help train and fine-tune foundational AI models.
The platform acts as a bridge between AI developers and human specialists, providing feedback that improves reasoning, accuracy, and contextual understanding. Mercor charges clients a finder’s fee and hourly rate for expert contributions, with contractors earning an average of $85 per hour.
The company now processes over $1.5 million in contractor payments daily and boasts a network of more than 30,000 active experts, supporting major AI labs worldwide.
Advertisement
Riding the AI Infrastructure Wave
The company’s rapid growth coincides with a shifting landscape in the AI training industry. Following reports that OpenAI and Google DeepMind reduced reliance on data-labeling vendor Scale AI, Mercor’s hybrid model — blending human expertise with reinforcement learning infrastructure — has become increasingly attractive.
The company has built software that supports reinforcement learning workflows, enabling human experts to validate or challenge AI-generated outputs in real time. Over time, these feedback loops improve model accuracy and performance, particularly in specialized or regulated fields.
“We’re helping AI models learn not just what can be done — but what should be done,” said a company spokesperson.
Scaling Toward Automation and Global Expansion
According to investor materials seen by TechCrunch, Mercor told backers it’s on track to reach $500 million in annual recurring revenue (ARR) faster than fellow AI unicorn Anysphere, which hit the same milestone roughly a year after launch.
With the new capital, Mercor plans to focus on three key areas:
- Expanding its global expert network across new disciplines and geographies.
- Enhancing its matching infrastructure to connect experts and clients more efficiently.
- Building automated tools to streamline human–AI collaboration and accelerate AI product deployment.
About Mercor
Mercor is a San Francisco–based AI infrastructure company connecting domain experts with leading AI labs to train and refine foundation models. Through a proprietary matching platform and reinforcement learning tools, Mercor enables human feedback at scale — helping AI systems learn nuance, intent, and judgment. Founded in 2022, the company works with thousands of professionals globally and has raised over $500 million to date from top-tier investors.
