Mercor
Vetted subject-matter experts for AI research, RLHF data, and AI-agent training.
About Mercor
Mercor connects AI labs and enterprises with subject-matter experts for frontier research, RLHF data production, and AI-agent training. Its Open Box model charges by the hour and applies a flat percentage to annotator pay, avoiding a fresh per-task negotiation for every iteration. Clients may work directly with annotators, know who handles their data, and keep work on their own platform. Mercor takes care of sourcing, vetting, and day-to-day expert management. Candidate evaluation can draw on interviews, work history, education, GitHub profiles, and Google Scholar citations. This setup fits teams that need specialized human judgment but do not want to build an entire recruiting and contractor-operations function. Every engagement starts with a custom proposal for scope and methodology, so buyers need to discuss requirements before estimating total cost or comparing the service with a standardized package.
Best for
Best use cases
Producing RLHF datasets with expert review
Training and evaluating domain-specific AI agents
Running research projects that require specialized judgment
Compare before choosing
More client visibility than a black-box data vendor
Less internal sourcing work than building an expert pool alone
More scoping effort than a self-serve software subscription
Features
Expert sourcing and vetting
Mercor evaluates candidates using interviews, work history, education, GitHub profiles, Google Scholar citations, and other relevant signals.
Open Box delivery
Clients can work directly with annotators, see annotator pay, and keep data on their own platform when preferred.
Managed expert operations
Mercor handles sourcing and annotator management while helping teams set up data pipelines and working processes.
Pros
- Hourly pricing avoids a new per-task price negotiation for every project iteration.
- Clients can keep work on their own platform and retain control of data and IP.
- Mercor handles expert sourcing, vetting, and management, reducing contractor-operations work.
Cons
- Custom proposals make total cost unavailable until scope and methodology are discussed.
- The Open Box workflow requires more direct client involvement with annotators than a fully outsourced vendor model.
- Hourly pricing can make final spend depend on project duration and expert pay rates.
Pricing Plans
Custom Open Box engagement
- Custom proposal based on scope and methodology
- Flat percentage of annotator pay rates
- Expert sourcing, vetting, and management
- Option to work on the client’s own platform
* Pricing information is for reference only. Please refer to the latest announcement on Mercor's official website.