Mercor Eyes Doubling in Valuation as Revenue Surges Past $2 Billion

A Rapid Ascent in the AI Infrastructure Layer

Mercor, the AI training platform that has become a critical piece of infrastructure for companies building autonomous agents, is negotiating a new funding round that would value the business at $20 billion, according to people familiar with the matter. The figure represents a 100 percent jump from the company's $10 billion valuation secured during its $350 million Series C in October.

The conversations remain at an early stage, but Mercor has already told potential investors it received a term sheet at the new valuation, according to sources. For a company operating in the notoriously capital-intensive AI training space, the speed of the valuation climb is notable - though not unprecedented in an environment where compute and training expertise command premium multiples.

At DailyTechWire, we've tracked dozens of AI infrastructure plays across Asia and North America over the past eighteen months, and Mercor's trajectory mirrors a broader pattern: platforms that can demonstrate both revenue scale and technical defensibility are finding themselves in bidding wars among late-stage investors eager to lock in exposure to the picks-and-shovels layer of the AI stack.

Revenue Velocity as the New Signal

The valuation talks arrive on the heels of a striking revenue disclosure. Brendan Foody, Mercor's founder and CEO, posted on X that the company's annualized revenue run rate has crossed $2 billion - a doubling in just four months. That pace of expansion is rare even among high-growth SaaS companies, and it suggests Mercor has found product-market fit with enterprises that need to train custom AI agents at scale.

The revenue acceleration also provides a clearer lens into why investors might be willing to underwrite a $20 billion valuation. At that figure, Mercor would trade at a 10× multiple on its current run rate - a premium, but not outlandish for a company growing at triple-digit rates in a category where switching costs are high and the addressable market remains vast.

What's driving the growth? Mercor's core offering sits at the intersection of human-in-the-loop training and automated feedback systems, a design that allows clients to fine-tune large language models and agent behavior without building in-house labeling and evaluation infrastructure. As more enterprises move from experimentation to production deployments of AI agents, the need for continuous training and quality assurance has become a bottleneck - one Mercor is well-positioned to address.

The Deeptune Acquisition and Vertical Integration

Mercor announced on Thursday that it is acquiring Deeptune, a startup specializing in AI agent training workflows. The entire Deeptune team will join Mercor as part of the deal, according to the company. Financial terms were not disclosed, but the move signals a strategic push toward vertical integration.

Deeptune's tooling focuses on reinforcement learning from human feedback and synthetic data generation - techniques that have become essential as the industry moves beyond static datasets toward dynamic, task-specific training loops. By bringing that capability in-house, Mercor can offer clients an end-to-end stack: from initial model selection and prompt engineering through iterative fine-tuning and deployment monitoring.

The acquisition also reflects a broader trend in the AI infrastructure market. As the technology matures, platform players are racing to own more of the value chain, reducing dependencies on third-party vendors and capturing margin that would otherwise leak to point solutions. For Mercor, the Deeptune deal is a bet that bundling training orchestration, evaluation, and reinforcement learning under one roof will create enough lock-in to justify the valuation investors are contemplating.

Putting Early Troubles in the Rearview

The revenue surge and acquisition talk stand in sharp contrast to the turbulence Mercor faced earlier this year. The company disclosed a data breach in the first quarter, an incident that exposed customer training datasets and prompted an internal security overhaul. Around the same time, several contract workers filed lawsuits alleging wage violations and misclassification - claims that drew scrutiny from labor regulators and briefly spooked some investors.

At the time, the combination of operational and legal headwinds led some observers to question whether Mercor's rapid scaling had outpaced its governance and compliance infrastructure. The company responded by hiring a chief information security officer, implementing stricter access controls, and settling several of the labor disputes out of court.

The fact that Mercor is now fielding inbound interest at double its previous valuation suggests the market has largely moved past those concerns - or at least decided that the company's growth trajectory outweighs the execution risk. That calculus may shift if new issues emerge, but for now, the narrative has pivoted back to revenue and product momentum.

The Competitive Landscape and What Comes Next

Mercor operates in a crowded field. Startups like Scale AI, Labelbox, and Snorkel have built significant businesses around data labeling and model training, while cloud incumbents - AWS, Google Cloud, and Microsoft Azure - are bundling training tools into their platform offerings. In Asia, companies like Alibaba Cloud's PAI and Tencent's TI-ONE are carving out regional footholds, particularly among enterprises wary of sending training data across borders.

What differentiates Mercor, at least according to the pitch it has made to investors, is its focus on agent-specific workflows rather than general-purpose model training. As AI moves from chatbots to autonomous systems that can book travel, negotiate contracts, or manage supply chains, the training paradigm shifts from maximizing next-token prediction accuracy to optimizing multi-step task completion and safety constraints. Mercor's infrastructure is purpose-built for that use case, which may explain why it has been able to command such rapid revenue growth.

The question now is whether the company can sustain that velocity. Doubling revenue every four months is not a steady state - it's a land-grab phase, and the easy customers have likely already signed. The next cohort will be harder to win, more price-sensitive, and more demanding on integration and support. If Mercor can navigate that transition while maintaining unit economics, the $20 billion valuation will look prescient. If growth decelerates sharply or burn rates spiral, the round could come to be seen as the high-water mark of AI infrastructure exuberance.

For now, the company appears to be leaning into expansion. The Deeptune acquisition adds engineering talent and product surface area, and the reported term sheet gives Foody the option to pour capital into sales, compute infrastructure, and international expansion. Whether that capital gets deployed wisely will determine if Mercor becomes a foundational layer of the AI stack - or another cautionary tale of a company that scaled too fast and stumbled on execution.