How long are dry spells between paid work on platforms like Outlier, Mercor, or DataAnnotation?

Work availability varies widely across platforms and specialisms. Our survey tracks what percentage of trainers who passed a platform's assessment have no work right now, and the longest stretch respondents went without paid tasks in the past six months.

Is the survey anonymous?

Yes. No login, no name, and your answers are never tied to your email. The optional discount code is captured separately with its own consent checkbox and is never joined back to your answers.

Community data

What do AI training platforms actually pay?

Q: How much do AI training platforms actually pay per hour?

Advertised rates and real effective hourly earnings are often different. Our survey collects active-week earnings divided by all-in hours, including unpaid screening, queue-watching, and rejected tasks. Results for each platform are published once enough respondents have contributed.

Advertised rates tell you what a task is worth. They don't tell you how much of your time goes unpaid, how long you wait between projects, or what you're actually earning per hour once you account for all of it. This survey measures that, from people doing the work right now.

AI training platform experience

Open now

Pay, work availability, onboarding, and support, from people on Outlier, Mercor, DataAnnotation, Alignerr, Micro1, Remotasks, and more. Anonymous, about two minutes, no login.

Eligible answers get a discount code for the certified exam, as a thank you.

Take the survey

What the data covers

Results for each platform are published once enough people have contributed. Here's what the survey measures.

True hourly rate

Active-week earnings divided by all-in hours, including unpaid screening, queue-watching, and rejected work. Not the task rate. What you actually earn.

Perceived vs. true gap

What trainers think they earn per hour versus what the math says. The gap is often wider than people expect once unpaid time is counted.

Work availability right now

What share of people who passed a platform's assessment have no paid work at this moment. A point-in-time fact, not a recall estimate.

Dry-spell length

The longest stretch without paid tasks in the past six months. Separates platforms that occasionally go quiet from those that go dark for months.

Platforms covered

Outlier Mercor DataAnnotation Remotasks Alignerr Micro1 Handshake Toloka Appen Scale AI Invisible AI Surge

Questions

Why can't I just look this up?

Platform websites list task rates, not effective earnings. Reddit threads are anecdotes. Neither tells you what a typical week actually pays once you count the screening calls, the waiting, and the tasks that got rejected. This survey asks people to report a real week's earnings and hours, and we compute the rest.

Is it really anonymous?

Your answers are never tied to your name, account, or email. The optional exam discount is a separate step with its own consent checkbox, and nothing from your answers travels with it.

How do I know the numbers are real?

We only publish figures once enough people have contributed for a platform, and every published stat shows the sample size it came from. We trim outliers before computing averages, and we never report a number without the context to read it honestly.

What do you do with the data?

Responses are pooled into aggregate statistics, never published individually. We use them to build honest platform summaries on this site. We don't sell the data and we don't share it with any platform.