Remotasks Review 2026: Is Remotasks Legit, or Just Still Online?

Remotasks is not shut down. Checked on 26 August 2026: the site loads, signups are open, and the official status page reports all four systems operational with 100% uptime over the previous 90 days and no incidents logged. It is owned by Scale AI, it pays weekly through PayPal and AirTM, and its homepage advertises availability in 90 or more countries.

That is the answer to the question most people are searching. The more useful finding is what turned up while verifying it. The FAQ section on the Remotasks homepage still contains Lorem ipsum placeholder text. A live marketing page for a platform backed by a multibillion-dollar company, shipping filler copy nobody has replaced.

That detail is the whole story in miniature. Remotasks is legitimate and running. It is also the part of Scale AI's business that stopped getting attention, because Scale moved its expert work to Outlier and left Remotasks as the low-rate, high-volume, geography-priced tier. Nothing is broken. Nothing is being invested in either.

This Remotasks review covers the verified operating status, why so many people think it closed, what it pays now, how it differs from Outlier, and seven alternatives that pay more for the same hours.

Is Remotasks legit? The 26 August 2026 status check

Here is what verification actually returned, rather than a recycled answer.

• The platform is live. remotasks.com returns a working homepage with an open signup flow.
• The status page is green. status.remotasks.com reports Dashboard, Tasking Portal, Training Center and Task Log all operational, with 100% uptime across the previous 90 days and no incidents.
• It has a real corporate parent. Remotasks is Scale AI's annotation platform. Scale sells training data to the largest AI labs.
• Money flows one direction. No signup fee, no equipment purchase, no paid training. Every fake remote-work scheme fails this test first.
• Payment is weekly. The homepage states taskers are paid weekly via PayPal or AirTM.
• It claims 90 or more countries, which is wider coverage than any expert marketplace offers.

And the caveat that matters:

• The homepage FAQ is unfinished. Lorem ipsum placeholder text sits in the live page where answers should be.

Legitimate and maintained are different claims. Remotasks clears the first easily. It does not clear the second, and that gap is worth understanding before you invest unpaid hours in its training academy.

Why everyone thinks Remotasks shut down

The belief is not irrational. It comes from three real events.

1. The 2024 country exits. Remotasks ended operations in Kenya, Nigeria and Pakistan effective 8 March 2024. Thousands of workers lost access, in many cases without a clear warning, and reporting at the time described whole countries being dropped without explanation. Scale later said the shutdown notice failed to reach Kenyan workers because of an internal error, and that all work completed under its guidelines was paid. For the people affected, the platform had effectively shut down, and they said so loudly enough that the impression stuck globally.

2. The Philippines reporting. A 2023 Washington Post investigation documented delayed and withheld payments to Philippines-based Remotasks workers. Separately, Nigerian contributors described completing more than twenty hours of Igbo language work across five projects for under a dollar. Those accounts circulate as evidence the platform is fake. They are better read as evidence about rates, dispute handling and how little bargaining power a tasker has, which is a different and more accurate complaint.

3. Scale moved the good work to Outlier. Contributors who joined for well-paid model evaluation found those projects appearing on Outlier instead. Some locations opening the Remotasks signup page are now shown a message pointing them to Outlier. When the interesting work leaves a platform, it feels closed even while the servers stay up.

Add the 2025 contraction across everything Scale owns, after Meta bought roughly 49% of the company and OpenAI and Google pulled back their contracts, and you get a platform that is online, funded, and much quieter than it was.

Remotasks vs Outlier: same company, two different deals

This is the single most useful thing to understand, and most reviews skip it.

• Remotasks is the volume annotation platform. Image labelling, bounding boxes, transcription, basic categorisation, plus a free training academy. Low entry bar, no credential requirement, rates set by geography, wide country coverage.
• Outlier is the expert marketplace. Rates are published by domain, and coding, math, legal and medical review pay several times the Remotasks band. It requires identity verification, usually a degree, and an unpaid subject assessment.

They are both Scale AI. If you have a degree, a professional credential or the ability to code, signing up for Remotasks instead of Outlier means accepting a fraction of what the same company would pay you. Our Outlier AI review covers that platform's rates, requirements and its own 2026 volume problem in full.

How Remotasks actually works

Unlike the expert platforms, Remotasks does not screen you once and let you in. It gates every task type separately, and you unlock them yourself.

1. Create an account and take a placement test. Basic details, then an assessment that sets your starting point and which categories suit you.
2. Take courses in the Training Center. Each task type has its own course and its own qualification exam. Passing the exam is what puts that task type in your queue. Nothing is unlocked by default.
3. Practice before you earn. Practice tasks build accuracy first. Paid assignments come once you meet the performance standard for that category.
4. Take a bootcamp for the hard categories. LiDAR annotation and semantic segmentation, the highest-paying task types on the platform, run intensive bootcamps of roughly 10 days to two weeks.
5. Build a quality score. Accuracy on completed work governs which tasks you can reach and how often your work gets rejected. It is the closest thing to a career here.
6. Progress from Attempter to Reviewer. New taskers start as Attempters. Consistently accurate taskers can become Reviewers, checking other people's work.

All of that training time is unpaid. Two weeks of LiDAR bootcamp is a real investment against a rate that is still capped by your country's band, and that calculation is the honest test of whether Remotasks is worth it for you.

What Remotasks pays in 2026

Two variables set your rate: the task category, and where you live.

• General annotation (image labelling, bounding boxes, basic categorisation): roughly $5 to $15 an hour
• Model response evaluation and RLHF work: roughly $15 to $40 an hour
• Specialist categories: roughly $20 to $50 an hour

Geography does more work here than on any other platform in this market. Remotasks recruits heavily in Southeast Asia, Africa and Latin America and prices accordingly, which is precisely why it looks like a strong option in some countries and an insulting one in others.

Three mechanics to know:

1. Pay is per task, not per hour actually worked. Your effective hourly depends on how fast you clear the queue, so early weeks on an unfamiliar task type pay badly by definition.
2. Training is unpaid. The academy courses that unlock task categories cost you time and return nothing directly.
3. Task supply is project-based. Which queues are open when you log in decides your week. This is the same structural problem every Scale platform has, and it got worse across 2025.

Should you sign up for Remotasks in 2026?

It still makes sense if:

• You are in a country where the Remotasks band genuinely competes with your local alternatives.
• You have no degree or professional credential, and the expert marketplaces would screen you out.
• You want the free training academy to learn annotation work before applying somewhere that pays more.
• You want a second queue for the weeks your main platform is quiet.

Skip it if:

• You can code. Outlier's coding tracks, micro1 and Alignerr all pay multiples of the Remotasks rate for work you can already do.
• You hold a professional credential in law, medicine, finance or a science. You are the customer these platforms are competing for, and Remotasks is not where that money is.
• You speak a language other than English natively. Language work is priced far better on the expert platforms than on volume annotation queues.
• You need a predictable weekly figure. No platform in this market provides one, and this one least of all.

7 Remotasks alternatives that pay more

Each one is covered in full in the best AI training jobs roundup, ranked with pay and entry bar side by side.

1. Outlier. Scale AI's own expert marketplace, so the closest thing to a direct upgrade. US contributors see roughly $15 to $30 an hour for general rating, $30 to $50 for coding and math, and up to around $100 for legal and medical review. Same parent, same weekly payouts, different tier of work. Volume swings with Scale's client contracts.

2. DataAnnotation. The lowest entry bar of the expert-adjacent platforms and the fastest onboarding. Base work starts around $20 an hour with the coding queue at $25 to $40. One unpaid assessment, one attempt, no retakes. The DataAnnotation review covers how that gate is scored.

3. Mercor. Publishes an hourly rate on nearly every listing, which is rare in this market, and advertises a $112 average. The Mercor review covers its narrow matching funnel. Rates track scarcity of expertise, so expert contracts in medicine, law and quantitative finance sit at the top of the range. Screens with an interview rather than a written test.

4. micro1. AI-vetted talent network with rates published per role. Engineering listings cluster around $40 to $60 an hour, and senior, medical, legal and PhD-level roles post $100 to $175. Screening is skill-based rather than credential-based, which suits engineers without a graduate degree. The micro1 review covers its proctored AI interview.

5. Alignerr. Labelbox's expert network, weighted toward coding and language work. Generalist projects post in the teens to twenties per hour, expert projects from the $30s up, with advertised specialist ceilings around $150. Check which payment model a project uses first, as the Alignerr review explains.

6. Terac. Paid research studies for verified professionals, priced per approved session rather than by the hour. Sessions commonly post between $100 and $350, which for a sub-hour interview outruns nearly every hourly rate in this market.

7. TELUS Digital. The closest like-for-like alternative if you want rater work rather than expert work: search quality rating, ads evaluation and AI response review, with structured onboarding and steadier project volume than Remotasks. Around $17 an hour for US raters, country-indexed elsewhere. Appen and Clickworker, now part of LXT, sit in the same tier at roughly $8 to $15 an hour.

How to pick, in one pass

• You hold a professional credential. Mercor first, then Alignerr. Terac if you will do interview-format research sessions.
• You can code. micro1 and Outlier's coding track. DataAnnotation's coding queue as a third.
• Strong English, no specialist credential. DataAnnotation, then Outlier's general rating work.
• A language other than English. Outlier and Alignerr both price language work properly. Do not accept an annotation-tier rate for it.
• No degree, and you want to start today. Remotasks or TELUS Digital, while you qualify somewhere better.

The pattern behind every platform on this list

Remotasks did not get worse because it stopped caring about you. It got quieter because the contracts moved, and Scale put its attention on the platform that serves them. Outlier is now learning the same lesson from the other end, with verified contributors sitting idle because a customer three layers up signed a deal with Meta.

The conclusion is the same either way. A single queue is not an income, no matter how legitimate the company behind it is.

Contributors who earn steadily are qualified on two or three marketplaces at once, so a quiet week on one is covered by another. Doing that used to mean checking six boards with six sets of titles, pay conventions and signup flows. Skillora's AI training jobs board collects live listings from Mercor, micro1, Alignerr, Outlier, DataAnnotation and Terac into one feed, refreshed daily, filterable by pay, skill and experience level. Every listing links back to the marketplace that posted it, you apply and get paid by them directly, and browsing is free.

One thing that catches Remotasks veterans off guard: the platforms paying three to ten times more mostly screen with a live or recorded interview rather than a task queue or a written test. If your experience is annotation work, that conversation is an unfamiliar skill, and it is the actual barrier between the two tiers. Running it once with feedback beforehand costs an evening.

Verdict

Remotasks is legitimate, operating, and paying weekly as of 26 August 2026. It is also the least valuable place to spend an hour in this market unless your country's band makes it competitive, and its unfinished homepage tells you how much attention it is getting from its owner.

Use it as a starting point or a filler queue, not a destination. Then qualify somewhere that pays for what you actually know.