Best Data Annotation Jobs

An independent guide to paid AI training work

Rankings

Best Data Annotation Jobs (2026)

Ten platforms that pay people to train AI, compared side by side. What they pay, who they want, how fast the money arrives and what workers actually run into.

By the Best Data Annotation Jobs team · · First published 8 November 2026 · How we make money

On this page
  1. The comparison
  2. Each platform in detail
  3. How we rank
  4. What about Remotasks?
  5. How to spot a fake data annotation job
  6. Questions
  7. Sources

DataAnnotationOutlierMercorAlignerrProlificAppen (CrowdGen)micro1TuringInvisible (Meridial)TELUS Digital AI

Data annotation work means teaching AI models by example. You might rate two chatbot answers, write a better one, check a model's maths, label images, or try to make a model say something it shouldn't. Most of it is remote, paid by the hour or by the task, and done as an independent contractor.

The catch is that every platform works differently. Some pay generalists, some only want people with a degree or a licence, and some go quiet for weeks between projects. This page puts the details in one place so you can pick where to apply first.

Every figure below comes from the platform's own pages or from reporting we link to in the sources. Where we could not confirm something, we say so. The "Facts checked" date above is the last time we rechecked them. Pay changes often, so check the platform's own page before you apply.

The comparison

Ten platforms compared on published pay, joining requirements, payout and verdict. Details and sources are in each platform's entry below.
PlatformPublished payTo joinPayoutBest forVerdict
1DataAnnotation$25 to $50/hr general; $40 to $150+ coding (starting rates)Fluent English, bachelor's or equivalent, assessment, ID checkPayPal, a few days after you request itWriters, generalists, coders, STEMLegit
2OutlierPer project. Official examples: up to $15 to $50/hr by role and countryAssociate degree+, ID, local phone, work visa, screeningsWeekly (Tuesdays): PayPal, Airtm or ACHGeneralists and subject expertsLegit, with caveats
3MercorPer project. Past pool rates about $40 to $180/hr by fieldAI interview, matching experience, supported countryWeekly: Stripe Connect, or Wise in some countriesWorking professionals and expertsLegit, with caveats
4AlignerrPer job card, e.g. $20 to $50 or $40 to $120/hrRecorded AI interview, ID check, supported countryWeekly, the Friday after: Stripe or PayPalExperts and researchers; some entry rolesLegit
5ProlificMinimum £6 / $8 an hour set by Prolific18+, verified account, supported countryPayPal from £6 / $6; first cash-outs wait 72 hrsBeginners doing short studiesLegit
6Appen (CrowdGen)Per project; expert ceilings up to $150 to $450/hrQualification step; rules vary by projectMonthly invoice; PayPal, bank and othersMultilingual contributors, expertsLegit, with caveats
7micro1Per project; no public rate cardAI interview, Deel contract, time trackingTwice a month via Deel (about 5th and 20th)Experts matched to client projectsLegit
8TuringNot published on its job postsVaries; e.g. 3+ years' experience, 20+ hrs a weekNot confirmed (contractor terms not public)Software engineersLegit
9Invisible (Meridial)Rate range shown on each engagementProfile, confirmation step, location rulesEvery two weeksDomain experts and studentsLegit
10TELUS Digital AIEstimated rate on each listing (not confirmed by us)Per role, country and language; exam and IDNot confirmedBilingual raters in their own countryLegit

Short version only. Each platform's full entry, with sources, is below the table.

Each platform in detail

01

DataAnnotation

Legit

Run by Surge AI (Surge Labs), a San Francisco AI data company founded in 2020.

Pay
From $25 to $50 an hour for general work, $40 to $150+ for coding, $40 to $125+ for STEM and licensed professional work (its own FAQ).
Best for
Strong English writers and generalists, plus coders, STEM graduates and licensed professionals.
Requirements
Fluent English, a bachelor's degree or equivalent experience, a starter assessment of about an hour, and a Persona ID check. Countries are not listed on its site.
Payout
PayPal. Deposits arrive within a few days of requesting them.
02

Outlier

Legit, with caveats

Operated by Scale AI (contracting entity Smart Ecosystem Inc., d/b/a Outlier AI). Also runs Remotasks.

Pay
Shown before each project; varies by skill, project and location. Official pages on 8 Nov 2026 showed examples from up to $15 an hour (English writing) to up to $50 an hour (biology; an agent-builder role in eight countries), and up to $27.50 an hour for coding in India.
Best for
Generalists and experts in coding, maths, STEM and languages.
Requirements
At least an associate degree, valid ID and a phone from your country, a visa that permits work, plus skill screenings.
Payout
Weekly, on Tuesdays, by PayPal, Airtm or ACH bank transfer.
03

Mercor

Legit, with caveats

Mercor.io Corporation, San Francisco, founded 2023. Backed by Benchmark, Felicis and General Catalyst.

Pay
Set per project. Mercor's standing expert pools (open networks, not specific vacancies) say past listings have paid about $40 to $75 an hour (maths) up to $60 to $180 an hour (medical). Specific openings show their own rate, such as $90 to $140 an hour for a US maths PhD role. Mercor cites an average of over $95 an hour.
Best for
Professionals and domain experts: software, medicine, law, maths, finance.
Requirements
A short AI interview (about 20 minutes on the listings we checked), experience that fits the role, and a home country on its supported list (108 when we checked). Its listings say it cannot support US H-1B or STEM OPT candidates.
Payout
Weekly through Stripe Connect, or Wise where Stripe is not available (its policy names countries such as Brazil, Kosovo and the UAE). The first Stripe payout has a 7-day hold, and banks can take several more days.
04

Alignerr

Legit

Alignerr LLC, powered by Labelbox, a San Francisco AI data company founded in 2018.

Pay
Ranges on each job card, for example $20 to $50 or $40 to $120 an hour. Its home page shows $80 an hour as average pay.
Best for
PhDs, researchers and experts, with some entry roles that need no experience.
Requirements
A recorded AI interview, a Persona ID check once matched to a project, and a supported country (not all countries can join).
Payout
Weekly. Stripe or PayPal depending on your country, paid the Friday after each Monday to Sunday week.
05

Prolific

Legit

Prolific Academic Ltd, London, UK.

Pay
Researchers must pay at least £6 / $8 an hour and are told to pay £9 / $12. AI tasks for invited taskers usually pay more.
Best for
Beginners who want short paid studies, and invited AI taskers.
Requirements
Age 18+, a verified account and residence in a supported country (mostly OECD, listed on its help centre). Unsupported countries can join a waitlist.
Payout
PayPal, once a day, from £6 / $6. The first four cash-outs wait 72 hours.
06

Appen (CrowdGen)

Legit, with caveats

Appen Limited, listed in Australia (ASX: APX), founded 1996. Contributor work runs on CrowdGen.

Pay
Per project. Its home page lists expert ceilings such as up to $150 an hour for software engineers; general data tasks pay per unit.
Best for
Multilingual contributors for data collection and rating, and domain experts.
Requirements
A qualification step for most projects, one account per person, and location rules that differ by project.
Payout
Monthly invoice finalised by the 14th. PayPal, local bank, Airtm, SWIFT, Payoneer or gift card, depending on country.
07

micro1

Legit

micro1 Inc., founded 2022. Raised $35M at a $500M valuation in September 2025.

Pay
Hourly or per task, set by project. No rate card on its own site; listings load per role.
Best for
Experts and professionals matched to client projects.
Requirements
An on-demand AI interview with its recruiter Zara, a contract through Deel, and Hubstaff time tracking for hourly work.
Payout
Twice a month through Deel, around the 20th and the 5th.
08

Turing

Legit

Turing Enterprises, Inc., Palo Alto, founded 2018. Valued at $2.2B in March 2025.

Pay
Not published. Its job posts list contract length and hours but no rate.
Best for
Software engineers and technical specialists.
Requirements
Varies by project, for example 3+ years of experience and 20 to 40 hours a week with 4 hours overlapping US Pacific time.
Payout
Not confirmed. Turing's public help article on payment dates and bank transfers covers referral rewards, not contractor pay, and its contractor payment articles need a login.
09

Invisible (Meridial)

Legit

Invisible Technologies Inc. Expert work runs through its Meridial marketplace. Valued at $2B in September 2025.

Pay
Set per project. Meridial says it publishes the rate range on each engagement before you commit; there is no general rate card.
Best for
Domain experts in finance, law, coding, STEM and languages, plus students.
Requirements
A profile, a short confirmation step, and project-by-project location rules.
Payout
Every two weeks to the account in your profile.
10

TELUS Digital AI

Legit

TELUS Digital (formerly TELUS International), owned by Canadian telecom TELUS since its October 2025 buyout.

Pay
Each listing shows an estimated hourly rate, paid per task. Its job pages block automated checks, so we have not confirmed current figures.
Best for
Bilingual raters for search, ads and maps in their own country.
Requirements
Apply per role, tied to your country and language, then pass a qualification exam with ID verification.
Payout
Not stated on a page we could open.

How we rank

We score each platform on five things, in this order:

  1. Pay you can verify. Rates published by the platform or in its own job posts count. Rumours do not.
  2. Getting paid. How you are paid, how often, and whether workers report late or missing payments.
  3. Steady work. Whether projects keep coming after you pass the tests, or dry up.
  4. Who can join. Countries, degrees and tests. A platform with lower published entry requirements ranks higher for beginners.
  5. Track record. Who runs it, how long it has been around, and any lawsuits or public problems.

Referral fees never change the order. See About and disclosure for how we make money.

What about Remotasks?

Remotasks is Scale AI's older tasking brand, and Scale also runs Outlier. Its website is still live and says it pays weekly by PayPal or AirTM, but in 2024 it closed in Kenya, Nigeria and Pakistan and paused sign-ups in several other countries. Wikipedia describes the two as separate Scale AI arms: Outlier for training chatbots and other language models, Remotasks for image, video and self-driving data. That is why we cover it inside Outlier's entry rather than ranking it separately.

How to spot a fake data annotation job

Scammers copy the names of real platforms. A few rules hold for every legitimate platform we track:

  • You never pay to join, to take a test or to "unlock" tasks.
  • Offers come from the platform's own website or email domain, not a personal chat app.
  • No one asks for your bank login, a crypto deposit or a fee for "training materials".
  • Real assessments happen inside the platform's own site after you sign up there directly.

If a recruiter contacts you first, go to the platform's website yourself and apply from there.

Questions

Which data annotation job pays the most?

On published figures, the expert platforms. DataAnnotation's FAQ lists coding work from $40 to $150+ an hour. Mercor's medical expert pool says past listings posted $60 to $180 an hour, and Appen's CrowdGen lists ceilings of up to $450 an hour for medical clinicians. Those top figures are for qualified experts, and "up to" means the ceiling, not a typical rate. For general work, DataAnnotation publishes $25 to $50 an hour.

Can I do data annotation with no experience?

Yes, on some platforms. General tasks such as rating or writing responses can need only strong written English and a passed assessment. Expert projects in coding, medicine, law and maths ask for a degree or professional experience. Prolific asks for no qualifications beyond being 18, verified and living in a supported country. Some Alignerr listings say no experience is needed. DataAnnotation asks for a bachelor's degree or equivalent experience, and Outlier for at least an associate degree.

Are data annotation jobs legit?

The ten platforms above are real companies that pay workers, and the reasons are linked in the sources. That does not make the work easy. Some reviewers report gaps between projects, accounts closed with little explanation and unpaid assessment time; the legit checks for Outlier and DataAnnotation link dated examples. Scammers also use real platform names, so always apply on the official website.

Is data annotation full time work?

Usually not. Almost every platform hires independent contractors with no guaranteed hours. Treat it as flexible side income unless a contract says otherwise.

How do data annotation jobs pay you?

It depends on the platform. Outlier, Mercor and Alignerr pay weekly, through PayPal, Airtm, bank transfer, Stripe or Wise depending on the platform. DataAnnotation pays by PayPal within a few days of your request. Prolific lets you cash out to PayPal once a day. micro1 pays twice a month through Deel, Invisible every two weeks, and Appen invoices monthly. Turing's contractor payment terms are not public, so we could not confirm them.

Sources

  1. Wikipedia: Scale AI (Outlier and Remotasks)
  2. Glassdoor review: Outlier deactivation (June 2026)
  3. Trustpilot review: DataAnnotation empty board (4 Oct 2026)
  4. Trustpilot review: CrowdGen pay shortfall (23 Oct 2026)
  5. DataAnnotation FAQs (pay tiers, requirements, assessment reply, PayPal), checked 8 Nov 2026
  6. DataAnnotation home page
  7. Clarkson Law Firm: class action against Surge AI (20 May 2025)
  8. PYMNTS: Surge AI fundraise report (July 2025)
  9. Outlier FAQ (operator, degree, ID, weekly pay methods)
  10. Outlier terms of use (contracting entities, effective 11 May 2026)
  11. Outlier coding page, India (up to $27.50/hr), checked 8 Nov 2026
  12. Outlier About page (role rate examples), checked 8 Nov 2026
  13. Outlier role: OpenClaw Agent Builder (up to $50/hr; AU, CA, FI, FR, IT, NO, UK, US), checked 8 Nov 2026
  14. Outlier community guidelines
  15. King & Siegel: Scale AI $12.5M misclassification settlement (Aug 2026)
  16. TechCrunch: second wage lawsuit against Scale AI (Jan 2025)
  17. Remotasks home page
  18. Rest of World: Remotasks closes in Kenya, Nigeria and Pakistan (Mar 2024)
  19. McKinney v. Scale AI settlement administrator FAQ (hearing date, defendants' denial), checked 8 Nov 2026
  20. PAGA filings list for Scale AI (30 Oct 2026 documents, posted 2 Nov 2026), checked 8 Nov 2026
  21. Mercor newsroom (founding, backers, average pay)
  22. Mercor supported countries
  23. Mercor time tracking and pay
  24. Mercor worker terms (updated 26 Aug 2026)
  25. Mercor listing: Expert Model Trainer
  26. TechCrunch: Mercor after its data breach (9 Apr 2026)
  27. Law360: AI startup misclassified 30K workers, suit says (11 May 2026)
  28. Futurism: Mercor project pay cut (Nov 2025)
  29. Mercor maths expert pool (standing listing, not a specific opening), checked 8 Nov 2026
  30. Mercor medical expert pool ("Healthcare Professional": past listings $60 to $180/hr), checked 8 Nov 2026
  31. Alignerr home page (pay ranges, average pay)
  32. Alignerr terms of use and applicant privacy notice
  33. Alignerr payment FAQs
  34. Alignerr Persona ID FAQs
  35. Alignerr: location denied error
  36. Alignerr: what is Alignerr
  37. Alignerr community rules
  38. Prolific: deciding on a reward (minimum and recommended pay)
  39. Prolific: AI taskers
  40. Prolific: who can take part (countries, age)
  41. Prolific: cashing out
  42. CrowdGen home page (operator, expert pay ceilings)
  43. CrowdGen payout options and fees
  44. CrowdGen invoicing
  45. CrowdGen contributor standards
  46. Appen: payments update from the CEO (Oct 2024)
  47. micro1 AI interview guide
  48. micro1 payroll policy
  49. micro1: working with Deel
  50. micro1 expert FAQs
  51. Dataconomy: micro1 raises $35M (Sept 2025)
  52. Turing Series E press release (Mar 2025)
  53. Turing help: how to get paid (Referral Program section; referral rewards only)
  54. Turing job post: LLM Trainer, SWE Bench
  55. TechCrunch: Turing raises $111M (6 Mar 2025)
  56. Meridial home page
  57. Meridial FAQ
  58. Business Wire: Invisible raises $100M (16 Sept 2025)
  59. TELUS completes privatization of TELUS Digital (31 Oct 2025)
  60. Wikipedia: TELUS Digital (rebrand, HQ)