Mercor lists more open roles than any other platform this site tracks. It hires professionals as remote contractors for AI training projects, screens almost entirely through a short AI interview, and pays weekly.
Every figure below comes from the roles this site checks each morning, or from Mercor’s own published policies.
The numbers
Mercor has 331 open roles carrying 12,370 stated openings.
Of those roles, 290 list an hourly rate, 36 pay per completed task, and 5 pay a fixed amount per session.
Among the hourly roles, the median advertised ceiling is $100 an hour. A quarter of them cap at $70 or below, a quarter at $120 or above, and the highest runs to $400.
Compare that with micro1, which lists 243 open roles but many times the openings, because a single micro1 listing can hire thousands of people for one annotation programme. Mercor roles are smaller and more specific: more listings, fewer seats behind each one.
The work is broader than “technical”
Mercor has a reputation for engineering and PhD-level work, and it does post plenty of that. But the spread across the roles open now is wider than the reputation suggests.
The largest groups are generalists (29 roles), business and operations (42), accountants and finance experts (34), bilingual speakers (26) and lawyers (23). Engineers account for 20 and software engineers for 18, with doctors and clinicians at 14 and life scientists at 14.
If you have a professional background in a white-collar field, there is probably something here aimed at it. The qualification is usually knowing a subject well, not knowing how to code.
Screening is one AI interview
Mercor’s published process is: upload a resume, complete whatever assessments the role requires, verify your work authorisation, then sit an AI interview.
Mercor says that interview takes about 20 minutes for most people and is the only way it screens candidates. There is no option to request a human interviewer. You can retake an interview up to three times.
Mercor also says it usually takes two to four weeks to hear back after applying. Applicants frequently describe waiting longer than that, with applications sitting in review well past the stated window.
One quirk worth knowing: some people receive offers for roles they never applied to. Completing interviews and assessments builds a profile that Mercor matches against incoming projects, so the screening you do for one role can surface you for another.
Hourly, per task, and why the difference matters
Most Mercor roles pay hourly, which its own documentation describes as the standard format. The 36 task-based roles work differently, and the difference is worth understanding before you accept one.
On a task-based contract you are paid for accepted output, not for time spent. Your submission goes to a reviewer, and if it comes back with corrections, the rework is unpaid. Contractors have raised a structural complaint about this: reviewers are typically paid hourly while the person doing the task is not, so the cost of another review round falls entirely on the task worker. Some also report that a resubmission can reach a different reviewer who flags a different set of issues.
We have not worked on these projects and cannot verify how common that is. But the asymmetry is real on its face, and it is worth weighing a task-based offer against an hourly one at a similar headline rate. An hour spent reworking a task pays nothing; an hour on an hourly contract pays the rate.
Some Mercor positions also start with a paid onboarding period of six hours.
Getting paid
Mercor pays weekly, processed on Wednesdays through Stripe. Where Stripe is unavailable, it uses Wise.
Weekly payment is faster than most of this market, and a real advantage over platforms running fortnightly or monthly cycles.
Where you need to be
Mercor sets eligible countries per role. Where a listing names no region, Mercor’s own policy says to assume it is open to candidates in the United States, Canada, the United Kingdom or the European Union.
Mercor is the only platform this site tracks that publishes a default that narrow. AfterQuery says it welcomes experts worldwide, and micro1 and Terac set eligibility per role without naming a fallback. So read the role’s own eligibility line before assuming you qualify.
The part people underestimate
Being accepted is not the same as having steady work, and the gap between the two is where most frustration with Mercor comes from. Applications sit in review. Projects end without much warning. Contract hour caps can be lower than you expect, so an accepted offer does not always mean a full week.
None of that is unique to Mercor, and it is covered in more detail in whether AI training can be a full-time income. The short version: treat a first Mercor contract as the thing that unlocks later work, because performing well on one project is what causes more to be routed to you.
Who Mercor suits
It is a reasonable fit if you have a professional credential or a specific domain background, can wait weeks for a decision without that being a problem, and want a platform that pays weekly once you are working.
It is a poorer fit if you need work immediately, if you are outside its default eligible regions and the roles you want name no others, or if you are looking for the high-volume generalist annotation work that micro1 carries more of.
Browse the Mercor roles open now, or read the micro1 and Mercor comparison if you are deciding between the two.