What if you could pitch your own AI research idea and receive funding, computing resources, and a substantial stipend to develop it? The Mercor Research Fellowship — APEX is designed for researchers, engineers, and academics working on AI evaluation, benchmarking, and economic research.
Quick Facts
- Opportunity Type: Research Fellowship
- Host Organization: Mercor
- Programme: Mercor Research Fellowship — APEX
- Location: Remote or in-person in San Francisco, with New York City and London also listed
- Mode: Remote / In-person
- Duration: 3 to 6 months
- Stipend: $40,000 for 3 months or $80,000 for 6 months
- Financial Support: Unlimited API credits, dedicated GPU compute budget, and paid expert/human-data support
- Nationality: Worldwide / no nationality restriction stated
- Degree Required: No specific degree requirement stated
- Preferred Background: Economics, Computer Science, Machine Learning, Statistics, Measurement, Psychometrics, HCI, Social Science, or related fields
- Age Limit: No specific age limit stated
- Work Experience: No minimum professional work experience requirement stated
- Work Commitment: Minimum 30 hours per week; full-time preferred
- Application Fee: $0
- Application Deadline: September 26, 2026, 9:00 AM GMT
- Application Requirement: CV, LinkedIn profile, and a one-page research proposal
- Nomination Required: No
- Research Focus: AI evaluation, benchmarking, evaluation methodology, AI economics, AI-for-science and related areas
What makes the APEX fellowship different?
Rather than following a fixed course or curriculum, applicants are expected to pitch their own research idea.
Your proposal can focus on developing the next generation of:
- AI evaluation benchmarks
- AI measurement frameworks
- Economic studies
- Research methodologies
- Evaluation systems for emerging AI models
There are also no nomination requirements. Applicants can apply directly and present their own research proposal.
Financial Support
Selected fellows receive significant financial and technical support:
- 3-month fellowship: $40,000 stipend
- 6-month fellowship: $80,000 stipend
- Unlimited API credits
- Dedicated GPU compute budgets
- Paid expert and human-data testing time
- Research resources and mentorship
Is there an application fee or hidden charge?
No application fee is specified.
According to the provided programme information, the fellowship is designed to provide funding and research resources rather than charge participants.
Who can apply?
The programme is aimed at people with a strong background in areas such as:
- Artificial intelligence
- Machine learning
- Computer science
- Economics
- Statistics
- Psychometrics
- AI evaluation and measurement
Applicants should also be prepared to commit at least 30 hours per week for the fellowship period.
Where can you work?
The fellowship offers flexible arrangements, with options including:
- Remote
- San Francisco
- New York City
- London
The exact arrangement may depend on the selected research project and fellowship terms.
How does the application work?
There is no nomination requirement.
The distinctive part of the process is that applicants pitch their own research or evaluation idea rather than simply applying to a predefined training programme.
A strong proposal should therefore clearly communicate what you want to investigate, why the problem matters, and what you intend to build or measure.
Application Deadline
September 26, 2026, at 9:00 AM GMT
If you’re interested in AI research, the interesting question isn’t just whether you can apply — it’s what would you build if you had $80,000, compute, APIs, and research support behind your idea?
Subscribe to Potenue for more opportunities that are worth exploring.