Signal

AI training startup Mercor raised a $750M Series D from Benchmark, General Catalyst, Nvidia, and others at a $20B valuation, up from $10B in October 2025

First reported by Reuters ·

The signal ●●●● Compiled by AI from Reuters, Techmeme, FinSMEs, Mercor, Mercor and 2 more
Why you might care

The cost to train AI models that perform complex professional tasks can now be measured in hundreds of millions of dollars.

What happened

AI training startup Mercor has secured $750 million in Series D funding, achieving a valuation of $20 billion. The funding round was led by investors including Benchmark, General Catalyst, and NVIDIA, with participation from numerous other firms and strategic partners like Salesforce and Zoom. Founded in 2023, Mercor utilizes a network of over five million human experts to train and evaluate AI models, enabling companies to refine AI performance on real-world professional workflows. The company plans to deploy this capital to enhance its research and development efforts, focusing on advanced evaluation techniques and realistic simulation environments for AI agent learning.

What it means

This significant funding round for Mercor underscores the immense capital flowing into AI infrastructure and specialized training solutions. The company's model, which relies on human expertise to refine AI, suggests a market maturation where raw AI model development is increasingly augmented by sophisticated post-training evaluation and alignment services. This focus on human-in-the-loop refinement at scale indicates a growing recognition that complex AI capabilities require more than just algorithmic advancement; they need rigorous validation against real-world professional standards.

The substantial valuation and backing from major tech players like Nvidia and Salesforce signal that investors are betting on specialized AI training and evaluation platforms as critical components of the AI ecosystem. This trend could lead to increased competition in the AI infrastructure space, with companies differentiating themselves through their ability to provide reliable, human-aligned AI performance. Businesses seeking to deploy advanced AI will likely face a landscape where third-party services for model validation and refinement become indispensable rather than optional.

AI-written summary. May contain errors.

AI Funding