Mercor
Mercor connects domain experts with AI labs for model evaluation and training work.
Mercor operates a dual-sided marketplace connecting domain experts (engineers, lawyers, consultants, doctors) with AI labs and enterprises for specialized evaluation, testing, and training work. The platform sources, vets, and places talent on project-based roles while AI companies use contractor feedback to refine frontier models. Mercor manages over 30,000 contractors paid collectively $1.5M daily, working with top-5 AI labs including OpenAI and Anthropic, and generates revenue through a margin on successful placements.
Problem solved
AI labs need domain-expert human feedback and evaluations to train and validate frontier models, but sourcing, vetting, and managing distributed expert contractors at scale is operationally complex.
Target customer
Frontier AI research labs and enterprise AI teams (OpenAI, Anthropic, and other top-5 AI labs; 6 of Mag 7 companies) seeking high-quality human feedback and domain-specific evaluations to refine AI models.
Founders
B
Brendan Foody
CEO & Co-Founder
Georgetown University dropout; founded Seros (cloud computing); Thiel Fellow; Forbes 30 Under 30 2025.
A
Adarsh Hiremath
CTO & Co-Founder
Thiel Fellow; college dropout; Forbes 30 Under 30 2025; former Bellarmine College Preparatory debate team member.
S
Surya Midha
Board Chairman & Co-Founder
Thiel Fellow; college dropout; Forbes 30 Under 30 2025; former Bellarmine College Preparatory debate team member.
Funding history
Seed
$3.6M
2023
Led by General Catalyst
· Soma Capital, Link Ventures, 2.12 Angels
Series A
$30M
September 2024
Led by Benchmark
· Jack Dorsey, Peter Thiel, Adam D'Angelo, Larry Summers
Series B
$100M
February 2025
Led by Felicis Ventures
· General Catalyst, DST Global, Benchmark, Menlo Ventures
Series C
$350M
October 2025
Led by Felicis Ventures
· Benchmark, General Catalyst, Robinhood Ventures
Total raised:
$492M
Pricing
Margin-based revenue model: Mercor applies a percentage fee to companies' pay rates once candidates accept roles. Early rates: ~30% margin on contractor wages (e.g., $500/week contract, contractor paid ~70%). Specific pricing not publicly disclosed; varies by expertise level and task complexity.
Notable customers
OpenAI, Anthropic, top-5 AI labs, 6 of Mag 7 companies
Tech stack
React (JavaScript frameworks)
Next.js (Web servers)
Webpack
Linkedin Insight Tag (Analytics)
reCAPTCHA (Security)
HSTS (Security)
Cloudflare (CDN)
Amazon S3 (CDN)
DoubleClick Floodlight (Advertising)
Amazon Web Services (PaaS)
Website
Competitors
Scale AI
Broader data-labeling and AI infrastructure platform; Meta invested $14.3B for 49% stake; serves wider use cases beyond model evaluation.
Labelbox
Data labeling and annotation platform; more focused on image/video labeling workflows; less specialized in expert knowledge work for AI model training.
Surge AI
Data labeling and AI training company; broader labeling services; less specialized in high-expertise domain work like legal, medical, financial consulting.
Why this matters: Mercor represents a novel model for training frontier AI: rather than building proprietary labeling infrastructure, it leverages distributed domain expertise to generate high-quality feedback that's difficult to automate. At $10B valuation after just 2 years and $492M raised, with three 22-year-old self-made billionaire founders, it signals massive investor conviction in the AI training/evaluation infrastructure market and the value of human expertise in AI development.
Best for: Best for frontier AI labs and enterprises that need rapid, large-scale access to domain experts (engineers, lawyers, doctors, consultants) to evaluate, test, and train frontier AI models with high-quality human feedback.
Use cases
Model Evaluation and Rubric Development
AI labs send evaluation tasks to domain experts—e.g., lawyers reviewing AI-generated legal summaries for accuracy or doctors checking medical reasoning. Experts provide detailed feedback that becomes training data to improve model performance. Mercor handles sourcing experts with the right credentials and managing the workflow.
Code Quality and Bug Detection
AI companies need engineers to review and fix code generated by AI models. Mercor's platform connects them with vetted software engineers who can identify bugs, suggest improvements, and provide feedback at scale. This loop trains models to generate better code.
Specialized Domain Feedback at Scale
Enterprises building AI tools for vertical markets (banking, healthcare, law) need domain experts to validate AI outputs. Mercor rapidly sources and deploys consultants, specialists, and practitioners to evaluate model performance against real-world requirements without hiring permanent staff.
Alternatives
Scale AI
Broader AI infrastructure and data-labeling platform with backing from Meta; serves more diverse use cases but less specialized in high-expertise consultant and professional feedback workflows.
Labelbox
General-purpose data labeling platform; primarily focused on image, video, and text annotation; lacks specialized recruitment and management of domain-expert professionals.
Surge AI
Data labeling and AI training service; broader labeling capabilities but not specialized in recruiting and deploying high-level consultants (lawyers, doctors, bankers) for expert evaluation.
FAQ
What does Mercor do? +
Mercor is a marketplace that connects domain experts (engineers, lawyers, doctors, consultants) with AI labs and enterprises for specialized work—evaluating AI models, testing outputs, and providing feedback to improve model training. Mercor handles sourcing, vetting, and payment of contractors, operating as a fully managed talent platform for AI-focused evaluation and training work.
How much does Mercor cost? +
Mercor uses a margin-based model: the company takes a percentage fee (approximately 30% based on early examples) from the contract value once a contractor is placed. Specific pricing varies by expertise level, task complexity, and contractor location. Contact Mercor directly for custom enterprise pricing.
What are alternatives to Mercor? +
Scale AI (broader AI data and infrastructure platform), Labelbox (general-purpose data labeling), and Surge AI (data labeling and AI training services). Scale AI is the closest competitor with significant venture backing (Meta investment) but serves a broader set of use cases beyond expert professional feedback.
Who uses Mercor? +
Frontier AI research labs including OpenAI and Anthropic, plus 6 of the Mag 7 companies. These organizations use Mercor to source domain experts for model evaluation, code review, and specialized feedback needed to train and validate cutting-edge AI systems.
How does Mercor compare to Scale AI? +
Mercor specializes in recruiting and deploying high-expertise professionals (lawyers, doctors, bankers, senior engineers) for expert evaluation and feedback; Scale AI is a broader AI infrastructure and data-labeling platform. Mercor focuses on expert knowledge work; Scale AI serves a wider range of labeling and annotation use cases. Scale AI has more capital ($14.3B Meta investment) but Mercor has achieved higher valuation growth ($10B in 2 years).
Tags
AI model evaluation
talent marketplace
data labeling
AI training
expert feedback
contractor management
frontier AI
model refinement