Last updated: 2026-06-20
Mercor connects 100,000+ domain experts with frontier AI labs for RLHF data and AI evaluation. Custom enterprise pricing; $10B valuation; $492M raised.
Mercor is an AI-powered talent marketplace founded in January 2023, valued at $10 billion after a $350 million Series C in October 2025. It connects 100,000+ domain experts (physicians, lawyers, PhDs, engineers) with frontier AI labs for RLHF training data and AI evaluation work. Pricing is custom enterprise only. Used by the top 5 AI labs and 6 of the Mag 7.
Mercor is an AI-powered talent marketplace founded in January 2023 in San Francisco by college dropouts Brendan Foody (CEO), Adarsh Hiremath (CTO), and Surya Midha (COO). The platform connects credentialed domain experts in medicine, law, investment banking, and software engineering with AI labs that need high-quality human feedback and training data for RLHF, alignment research, and AI agent evaluation. As of October 2025, Mercor raised a $350 million Series C at a $10 billion valuation, growing annual run-rate revenue from low double-digit millions to over $850 million in under two years. The platform's screening mechanism is a 20-minute AI-conducted video interview that generates real-time questions based on each candidate's resume and work history. Candidates who pass skip HR entirely and get matched directly to projects at partner labs. On the employer side, clients at team.mercor.com search across resumes, GitHub profiles, portfolios, and AI-generated interview transcripts using natural language. Mercor handles all five data modalities used in frontier AI training: text, image, audio, video, and code, covering every input type that top labs use in RLHF pipelines. Mercor is best suited for frontier AI labs running evaluation, red-teaming, synthetic data generation, and long-horizon agent testing that requires specialized professional knowledge. Its network of 100,000+ credentialed professionals in 190+ countries can complete tasks involving clinical reasoning, legal document review, financial modeling, and advanced coding that platforms relying on general crowdwork cannot match. Weekly direct payouts at $60 to $200 per hour make it one of the highest-paying AI training platforms available to contractors. In March 2026, Mercor launched Mercor Enterprise AI, expanding the platform to help enterprises build an organizational context graph from 34+ data source integrations including CRMs, ticketing systems, codebases, and communications tools. Pricing is custom enterprise with no public tiers; candidates apply and interview for free, while employers pay hourly rates negotiated through a sales process. The same month, a data breach exposed approximately 4TB of contractor data including video interview recordings and biometric data, leading to 7 class-action lawsuits and material reputational risk for prospective enterprise clients.
Custom enterprise pricing only; no public rate card. Candidates apply and interview free. Contractors earn $60-$200/hour; Mercor takes an undisclosed commission. Enterprise clients billed hourly per expert via an Experts-as-a-Service model. Demo required for employer pricing.
Mercor is an AI-powered talent marketplace founded in January 2023 in San Francisco, valued at $10 billion after raising $492 million in total funding, and used by the top 5 AI labs and 6 companies in the Mag 7. The platform connects 100,000+ credentialed domain experts in medicine, law, investment banking, and software engineering with AI companies that need high-quality human feedback and training data. Mercor's core product is a 20-minute AI-conducted video interview that screens candidates and matches them directly to RLHF, red-teaming, and AI agent evaluation projects. On the employer side, hiring managers search candidates by natural language across resumes, GitHub profiles, and AI-generated interview transcripts at team.mercor.com. The platform handles text, image, audio, video, and code modalities, covering every data type frontier labs use in alignment and training pipelines. In March 2026, Mercor also launched an Enterprise AI platform that builds an organizational context graph from 34+ data integrations to help companies optimize AI agent performance.
Mercor does not publish a public pricing page; all employer pricing is custom enterprise and requires contacting the sales team for a quote. The platform uses an Experts-as-a-Service billing model where enterprise clients pay hourly rates for each domain expert's time, with Mercor taking an undisclosed commission from contractor earnings. Contractors apply and complete the AI interview for free; if matched to a project, they earn between $60 and $200 per hour depending on domain expertise and project type. There is no free trial or self-serve tier for employers. Candidates receive weekly direct payments for completed work, with no reported minimum payout threshold. Budget planning for enterprise clients requires a full sales cycle before any costs are known.
Mercor's primary feature is an AI-conducted 20-minute video interview that generates questions in real time based on a candidate's resume, producing a scored transcript employers can review instantly without any HR involvement. The natural language search at team.mercor.com lets hiring managers query candidate profiles, GitHub repos, portfolios, and interview transcripts using plain English, removing Boolean search setup. Mercor's expert network covers 100,000+ vetted professionals in 190+ countries, including physicians, lawyers, PhDs, and investment bankers, enabling specialized RLHF tasks that general crowdwork platforms cannot fill. The platform handles all five AI training data modalities: text, image, audio, video, and code. Integrated payroll and compliance management handles cross-border payments in 190+ countries with weekly direct payouts to contractors. The March 2026 Mercor Enterprise AI addition builds an organizational context graph from 34+ data source integrations to help enterprises measure and improve AI agent performance.
Mercor is free for contractors: applying, completing the AI video interview, and building a profile costs nothing. If matched to a project, contractors are paid weekly at rates between $60 and $200 per hour; Mercor takes a commission from those earnings but does not publicly disclose the percentage. For employers and enterprise clients, there is no free tier: pricing is custom and requires a sales conversation. There is no self-serve dashboard employers can try without going through the sales process. Mercor does not offer a free pilot or proof-of-concept access. Prospective employers should budget for a minimum multi-month engagement given the custom onboarding and project scoping process.
Scale AI's Outlier platform is Mercor's closest direct competitor for RLHF and AI training data work, with more consistent task volume but lower typical pay rates of $20-$40 per hour versus Mercor's $60-$200 range for domain experts. Appen has operated in the human-in-the-loop data collection space longer and offers more task variety, but its pay rates and expert screening depth are lower. Toptal operates a similar expert-vetting model for software engineers, accepting only the top 3% of applicants, with a focus on long-term placements rather than AI training data. Surge AI and Labelbox both offer managed annotation pipelines for AI labs but focus on non-specialist data labeling rather than the credentialed professional expertise Mercor targets. For enterprises needing RLHF data at scale without a talent marketplace wrapper, contracting directly with annotation providers like Appen or CloudFactory may offer more pricing transparency.
Mercor is best for frontier AI labs and large technology companies that need specialized professional knowledge for RLHF, alignment research, or agent evaluation work that general crowdwork platforms cannot provide. The platform is ideal for AI research managers who need to quickly staff 10 to 500 credentialed experts for time-sensitive training data projects. On the contractor side, Mercor is well suited for physicians, lawyers, investment bankers, PhDs, and senior software engineers who want high-paying part-time remote work and can demonstrate 3+ years of verifiable domain expertise. Mercor is not suitable for contractors looking for a primary income source: project availability is inconsistent and there is no guarantee of work even after passing the AI interview. It is also not appropriate for heavily regulated enterprise clients who require verified SOC 2 or HIPAA compliance certifications before onboarding a data vendor, given the 2026 breach and compliance certification issues.
Yes. Mercor's interview process is conducted entirely by an AI system, not a human recruiter. The AI generates questions in real time based on the candidate's uploaded resume, LinkedIn profile, and work history, conducting a 20-minute video session where candidates respond verbally. After the interview, the AI produces a written transcript and a match score that hiring managers can review at team.mercor.com. Candidates are allowed up to 3 retakes if unsatisfied with their first attempt. The AI evaluates not just the content of answers but also communication clarity and domain-specific reasoning. Human review by the client company comes only after the AI has already scored and ranked candidates.
Mercor and Scale AI's Outlier both serve frontier AI labs with human feedback data, but they target different expert tiers. Mercor focuses on credentialed professionals (physicians, lawyers, bankers, PhDs) for specialized RLHF tasks and pays $60-$200 per hour, while Outlier serves a broader contractor pool with more diverse task types at lower typical rates of $20-$40 per hour. Mercor uses a proprietary AI video interview as the primary vetting mechanism; Outlier uses domain-specific written assessments and skill tests. Outlier offers more consistent task volume since Scale AI's lab relationships span more concurrent projects, while Mercor's work availability is more variable per contractor. For employers, Mercor emphasizes deep professional domain knowledge and custom project scoping; Outlier provides faster, higher-volume task deployment. The March 2026 Mercor data breach has not been matched by a comparable incident at Scale AI, giving Outlier an advantage in enterprise security conversations.