Mercor
Mercor · Coding · Business
Mercor is an AI data platform and expert data annotation service that connects frontier labs and enterprises with vetted experts who create training data.
About Mercor
What Is Mercor
Mercor is an AI data platform built around a network of human experts. Labs and companies need people who can grade a model's legal reasoning or write a realistic investment banking task. Mercor finds those people, manages the work, and pays them. The company started in 2023 as a hiring tool and pivoted once it saw how much frontier labs would pay for high quality expert data.
The business has three sides. Experts use Mercor to find paid remote work in fields like medicine, law, accounting, and software engineering. Labs use it to source human data and run evaluations at scale. Enterprises use it to turn their own workflow data into training material, or to deploy AI with help from Mercor's engineers.
The catch is who this fits. Mercor is a B2B service, not a self-serve tool you sign up for and use in an afternoon. Enterprises usually start with a sales conversation. Experts get screened before they see paid work, and approval depends on their field and experience. On the expert side, rates vary by task and project, so the hourly numbers on the job board aren't a guaranteed income.
Getting Started
- Apply at mercor.com as an expert or book a call as an enterprise. Experts record a short AI interview and build a profile.
- Add your specialty and evidence of it, like a license, portfolio, or work history. Matching depends on the field you qualify for.
- Get matched to projects. Labs and enterprises post tasks, and Mercor routes the ones that fit your verified skill.
- Complete the work in Mercor's platform, from writing prompts to grading model outputs.
- Get paid for approved work. Contracted experts are paid on a regular schedule through the platform.
Product Information
A quick look at Mercor's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- Domain experts hunting remote contract work
- Frontier AI labs
- Enterprise teams with valuable internal data
- Companies deploying AI that need help
Tasks
- Building training and evaluation data
- Running model benchmarks
- Creating reinforcement learning environments
- Monetizing internal workflow data
Scenarios
- A hospital network wants clinical data handled under strict privacy
- A lab needs 500 lawyers to grade contract analysis in a week
- A consulting firm wants AI that understands how it actually works
Key features
A vetted expert network
Mercor manages more than 30,000 contracted experts across medicine, law, finance, and engineering. Each one goes through an AI interview and a profile review before taking paid work. That screening is the core of the product, since labs pay for judgment that generic crowdworkers can't provide.
Human data and evaluation services
Labs come to Mercor for human data used in post-training, and for evaluations that check what a model can actually do. The platform works as both a data engine and an AI evaluation platform, and it runs these checks at a scale that matches how often frontier models ship. According to Mercor, the company pays its contractors more than $1.5 million a day.
The APEX benchmark family
APEX tests whether frontier models can do economically valuable work in professional services, accounting, and software engineering. APEX-Agents covers multi-hour tasks across investment banking, corporate law, and consulting. APEX-SWE measures real software engineering, and Mercor built it with Cognition. The company publishes the leaderboards and the methodology behind them.
Data monetization with anonymization
Enterprises can license the workflow data they already produce. Mercor's pipeline masks more than 60 categories of sensitive identifiers automatically, and the organization keeps control over what gets shared. That control matters for firms sitting on regulated data they can't just hand over.
Enterprise AI deployment
Mercor sends forward deployed engineers to help companies find high-value AI opportunities, map workflows, and ship production systems. The platform then measures performance and supports ongoing tuning. It's the services side of the business, aimed at buyers who want results rather than access.
Research and licensing partnerships
Mercor works with industry partners and its own experts to build the tasks behind every benchmark. It also offers license-ready datasets, expert-written and graded, with over 50,000 tasks across 30-plus domains. For teams that would rather download data than commission it, that's the path.
Pros and cons
Pros
- Deep expert screening, so labs get domain judgment instead of generic labels.
- Benchmarks like APEX-Agent and APEX-SWE measure real professional work, not trivia.
- Anonymization that masks 60-plus identifier categories supports data deals in regulated fields.
- The network scales hiring fast when a lab needs hundreds of specialists at once.
- Enterprise services cover the whole path from finding AI opportunities to deploying them.
Cons
- There's no free tier or self-serve signup, so small teams face a sales conversation before they can evaluate it.
- Expert pay varies by task and project, which makes income hard to predict month to month.
- Pricing for enterprise data and services isn't public, so buyers can't compare costs without contacting sales.
- Approval gates slow things down: both experts and enterprises wait on screening or scoping before work starts.
Frequently asked questions
Mercor connects AI labs and enterprises with vetted human experts. Those experts create training data, grade model outputs, and build evaluations and reinforcement learning environments for frontier AI models.
Related content
Explore related tools, skills, and articles for Mercor.
Mercor Alternatives
Base Code
Base44 (Wix) · Coding · ProductivityBase Code is a shared cloud development environment from Base44. Teammates describe changes to an existing GitHub codebase in chat, preview them live and send them back as pull requests.
Google EmbeddingGemma 2
Google DeepMind · Writing · Image · CodingGoogle EmbeddingGemma 2 is a 740M-parameter open model that turns text, code, images, audio, and video frames into vectors in one 768-dimension space, running on phones and laptops.
GitGlow
GitGlow · Coding · ProductivityGitGlow is a local-first desktop Git client that lets independent developers review their own code and their AI agents' changes in one place before shipping.