Published rates for remote annotation and AI training work run from about $10 per hour for general tasks on some platforms to $40–$150 or more per hour for expert work in fields such as law, medicine and software, according to the rates platforms publish themselves (as of October 2026). What you actually take home depends on how many hours of work are available, unpaid time, payout fees and taxes.
This article lists the published rates with their sources, explains what moves your rate up or down and shows the gap between an hourly rate and a monthly income.
Published Pay Rates
These are the ranges that large annotation and AI training platforms publish on their own websites. They are advertised ranges, not guarantees: actual rates depend on the project, your skills and often your country.
| Type of work | Published range |
|---|---|
| General tasks: labeling, rating, simple writing | about $10–$50 per hour |
| Language specialists | about $25–$40 per hour |
| Professional fields: law, finance, medicine, accounting | about $40–$125+ per hour |
| Coding, data and machine learning, engineering | about $40–$150+ per hour |
| Paid research studies and surveys (not annotation) | at least about $8 per hour; around $12 recommended |
| AnnoTrue, all projects | never below $10 per hour or the legal minimum where you live, whichever is higher; each project shows its rate before you apply |
Ranges checked on 06.10.2026 on the pay pages of several large platforms. They change often.
Two patterns are clear. First, expertise pays: professional and technical work is advertised at several times the rate of general tasks. Second, ranges are wide even for the same kind of work, so the project and your track record matter as much as the platform.
What Drives Your Rate
| Factor | Effect on pay |
|---|---|
| Expertise | Professional degrees and experience (law, medicine, software, finance) command the highest rates |
| Task type | Writing and evaluating AI answers usually pays more than simple tagging |
| Language | Native speakers of in-demand or less common languages can earn more on language projects |
| Location | Some platforms adjust rates by country |
| Quality record | Consistent high accuracy leads to more invitations, reviewer roles and better projects |
| Pay model | Hourly and per-task pay can differ a lot in practice, depending on how fast you work accurately |
| Role | Reviewers usually earn more than annotators on the same project |
Per hour or per task?
With hourly pay, you are paid for time worked, which is fairer when items vary in difficulty. With per-task pay, your effective hourly rate depends on your speed, which rises with practice, but rushing to raise it usually costs accuracy and accuracy is what keeps you on projects.
Hourly Rate vs. What You Actually Take Home
An hourly rate is not a monthly income. Four things sit between them:
- Available hours. Projects open and close; some weeks have plenty of work, others none.
- Unpaid time. Sign-up, assessments, some onboarding and reading guideline updates are often unpaid.
- Payout fees. Wise, Payoneer, PayPal and banks charge transfer or currency fees.
- Taxes and social contributions. As an independent contractor you usually declare and pay them yourself.
An illustrative month (example numbers, not a promise):
| Item | Amount |
|---|---|
| Paid hours worked | 60 hours |
| Rate | $15 per hour |
| Gross earnings | $900 |
| Unpaid time (assessment, onboarding, guideline reading) | 5 hours |
| Effective hourly rate including unpaid time | $900 ÷ 65 hours ≈ $13,85 |
| Payout fee | depends on the method you choose |
| Taxes and contributions | depend on your country and status |
The lesson: compare platforms on steady, paid hours at a fair rate, not just the top of an advertised range. A $50 headline rate with little available work can pay less than a $15 rate with steady projects.
How to Earn More
- Keep your accuracy high. It is the main thing that brings invitations to better projects and reviewer roles.
- Qualify in your field. If you have professional training, expert projects pay the most.
- Add languages. A strong second language opens language-specific projects.
- Move toward evaluation and writing tasks, which usually pay more than simple tagging.
- Choose your payout method wisely, comparing fees before you pick one.
- Work with more than one reputable company to smooth out quiet periods.
- Never pay to get work. Real companies pay you; see the scam signs in How to Become a Data Annotator.
Clear Pay, Shown Upfront
At AnnoTrue, every project shows its rate, hours and payment schedule before you apply and pay is never below $10 per hour or your local legal minimum. Payouts are monthly, by the method you choose, with fees shown before you choose.
About These Numbers
The ranges come from the pay pages that large annotation and AI training platforms publish themselves, checked on 06.10.2026. We don't name individual platforms because their rates change often.