AIPayList

Live comparison

Outlier queue empty? Explore other AI training jobs

If your Outlier queue is empty, use this page to compare other AI training platforms before starting another assessment. Check the work, advertised pay and entry requirements first. An open application does not guarantee a project or immediate tasks.

Outlier is the best-known name in paid AI training, and through 2025–26 it also became one of the most frustrating: workers widely report empty task queues, project pauses and rates assigned only after you're inside, while the public site advertises “up to” ceilings rather than actual project rates.

The good news: several platforms doing the same work (model evaluation, expert annotation, AI training) publish concrete rates on open, dated listings you can compare before spending hours on onboarding. Everything below is read live from those platforms' own feeds (3234 open roles tracked right now).

Why people leave Outlier

Three complaints dominate worker reports. First, queue drought: you pass onboarding, then “no tasks available” for days or weeks, income you can't plan around. Second, opaque pay: the advertised number is a ceiling (“earn up to $40/hr”), and your actual project rate (often far lower, and different per project) appears only once you're in. Third, unpaid gating: assessments and requalification exams take real hours and pay nothing, which stings twice when the project you qualified for pauses a week later.

None of this makes Outlier a scam. It makes it a lottery. The rational response is the one experienced AI workers already use: apply to two or three platforms that show their rates up front, and let Outlier be one queue among several instead of your single point of failure.

Public worker reports and the platform's own pages, gathered by hand · 2026-09-03 · medium confidence

  • ▲ The advertised number is a ceiling. The project rate appears after onboarding and can sit far below it.
  • ▲ Task queues can sit empty for weeks, which is what turns an advertised hour into a much smaller effective one.

This directory can show: Public project listings, advertised hourly ceilings and each role's allowed countries. General interest forms and conflicting pay are excluded. It cannot measure: Queue depth inside Outlier, or the project rate you will be assigned.

Sources: indeed.com (2026-08) · techcrunch.com (2025-01-09)

At a glance: live advertised rates per platform

PlatformMedian top rateTypical bandTop rateDistinct ratesRate before applyHow you get in
Mercor$105/hr$70–$150/hr$450/hr83 (215 listings)Published rates on listings this directory readsOpen listings you can apply to
Alignerr$75/hr$50–$120/hr$400/hr67 (627 listings)Published rates on listings this directory readsPublic job board with hourly bands
Meridial (Invisible)$45/hr$28–$65/hr$200/hr29 (762 listings)Published rates on listings this directory readsOpen freelance project listings
micro1$90/hr$45–$130/hr$400/hr128 (343 listings)Published rates on listings this directory readsApply once, then matched to projects
Outlier (for comparison)public board advertises "up to" ceilings, project rates are assigned after onboardingPublished rates on listings this directory readsApply to a listed project, then complete onboarding

Distinct advertised rates (one vote per platform + rate combination), percentiles of band tops, recomputed from live feeds. Methodology · full pay index · what happens after you apply

Where to apply instead, and who each one suits

Compare Outlier head to head

How to switch without losing income

Apply in parallel, not in sequence. Assessments and reviews take days on every platform, so serial applications waste weeks. Pick the two or three platforms above that match your profile and submit the same afternoon. Keep any Outlier queue you have: leaving is free, and stacked queues are the point.

Lead with your strongest credential. Expert platforms sort applicants hard: a specific license, degree, portfolio or GitHub moves you into the $100+/hr tier; a generic résumé lands in the generalist pool. If you speak a language other than English natively, say so everywhere. It's the single most underused qualifier in AI work.

For language work, compare AI training jobs by language across the platforms we track. Check each role's language and country requirements on its original listing before applying.

Track your effective hourly rate. A “$25/hr” project with slow task loading and unpaid review time can net less than a $18/hr project with a full queue. One week of honest time-tracking per platform tells you where your hours actually earn most.

Re-check the feed weekly. Roles and rates turn over constantly: the top of the live board looks different most weeks. The best-paying openings tend to fill fast, and new expert tracks appear with no announcement.

Frequently asked

Which Outlier alternative pays the most?

Mercor currently advertises the highest rates in our feed: top rates typically $70–$150/hr, the highest $450/hr, measured from 83 distinct advertised rates. Our highest-paying platforms ranking recomputes from live feeds continuously, so treat this answer as of today rather than fixed.

Is Outlier AI legit, or should I avoid it?

Outlier is legitimate. It's operated by Smart Ecosystem (a Scale AI company) and has paid out to a large contributor base. The frustration is operational, not fraud: task queues can sit empty for weeks, projects pause without notice, and your actual project rate is often only visible after onboarding. This page exists so you can compare platforms that publish concrete rates before you invest hours in assessments.

Can I work on several AI training platforms at the same time?

Usually yes, and experienced AI workers treat it as standard practice: queues on any single platform are volatile, so stacking two or three platforms smooths your income. Check each platform's terms for exclusivity clauses on specific projects (rare but they exist), and never reuse client-confidential material between platforms.

Do these platforms hire outside the United States?

Yes. That's one of the strongest reasons to look beyond Outlier. Meridial recruits across dozens of countries and languages, Mercor hires worldwide for many expert roles (though some are US-only), and micro1 runs globally too. Use our country pages to see exactly what's eligible where you live.

How long does onboarding take on the alternatives?

Mercor runs an AI interview plus profile review, micro1 screens through its own interview, and Meridial reviews applications per project. In my own experience a first Mercor offer took anywhere from a few days to a month, and Mercor automatically rejects an application that reaches one month without a decision, so a month of silence is a no. That is the practical difference from a silent waitlist: on Mercor you know where you stand within a month.

Are the rates on this page real or marketing numbers?

Every rate is read from the platform's own public job feed several times a day and shown as the top of its advertised band, the most the role says it can pay. Per-task pay is shown per-task, never converted into a fake hourly figure. The full rules are on our methodology page.

Looking for general evaluation or language work?

The highest advertised rates below skew to specialist expertise. If you came from Outlier's general rating queue, the live lists for model evaluation and multilingual work show every open role of that kind with its advertised rate, and the country pages show what is eligible where you live.

Highest advertised rates open right now

To be fair to Outlier: it still pays reliably for many contributors and its profile stays in our directory. This page exists because you deserve to compare real numbers first. Some links are referral links, always labeled, never a ranking factor (disclosure).