Data Labeling Jobs: Real Mercor and micro1 Openings (2026)

Real Data Labeling Openings on Mercor Right Now

Checked directly on Mercor's own job board this week, these are real, currently posted roles in the data labeling and annotation category.

Mercor has also posted dedicated "Data Labeling Specialist" and "Data Labeler" roles in the past, including a UK-based listing around $36/hr, though specific postings like these open and close as projects staff up, so don't expect every example here to still show as open the moment you check.

The pattern worth noting: Mercor's data labeling pay isn't one number, it's a spread from generalist work in the $23-60/hr range up to specialist, credential-gated tasks at $65-75 per task. Apply directly through Mercor to see what's currently open in your specific background.

Real Data Labeling Openings on micro1 Right Now

micro1 is running one of the largest single data labeling pushes in this category right now: a public call to add 10,000 contributors within seven days specifically to review and label videos of robots performing tasks, open to English speakers in Western countries, part of a broader robotics and human-activity video annotation project.

The $15-80/hr spread on the main Video Annotation Expert role reflects the same pattern Mercor shows: broad task categories pay less, specific judgment-heavy work on the same project pays more. The $7/hr listing is the outlier here and a useful reminder to confirm the actual rate on any specific task before you commit time to it, not just the headline range on the project.

Apply through micro1 to see current openings, since a push this size tends to fill quickly once it's posted publicly.

Why Data Labeling Pay Swings So Widely

The same platform, sometimes the same company's data labeling category, can show pay anywhere from $7 to $95 an hour, and the reason isn't inconsistency for its own sake, it's that data labeling covers wildly different levels of judgment required.

A real, documented example makes this concrete: Business Insider reported that when Mercor's labeling work for a Meta project wound down, some contributors were moved to a different project paying $16 an hour, a real drop from what they'd been earning. In the same reporting, Mercor's own CEO cited an average pay figure of $95 an hour across the platform, while several contributors said they'd personally earned up to $60 an hour on other Mercor projects. All three numbers, $16, $60, and $95, can be true at the same time, because they're describing different projects, not a single stable rate.

The practical takeaway: a data labeling platform's advertised average or up-to pay figure describes the ceiling, not what you'll earn on any specific task you're assigned. The projects with the highest judgment requirements, specialist knowledge, nuanced quality review, pay closest to that ceiling. High-volume, low-judgment microtasks pay closest to the floor, regardless of which platform is running them.

What Data Labeling Work Actually Involves

Data labeling covers a wider range of tasks than the name suggests, and the specific type you end up doing has a direct effect on pay.

• Image and video annotation: drawing bounding boxes, tagging objects, or reviewing robot-task videos frame by frame, the category micro1's current robotics push falls under
• Text and preference labeling: comparing two AI-generated responses and picking the better one, a common task type on both Mercor and micro1 that pays more when it requires subject-matter judgment
• Transcription and audio labeling: converting speech to text or tagging audio for tone, accuracy, or language-specific detail
• Specialist review and red-teaming: credentialed professionals (chemists, safety engineers, scientists) evaluating AI outputs in their specific field, the highest-paying category on this page at $65-75/task

Generalist roles, Mercor's $50-60/hr MacBook-only listing among them, typically combine several of these task types into one role rather than specializing in just one. Specialist roles are narrower but pay closer to the top of the range.

How to Apply and What to Expect

Both platforms follow a similar pattern: sign up, complete a qualification step, then get assigned to whatever project currently matches your profile.

Mercor's qualification stage is typically paid, onboarding and training time count as billable hours on most of its client projects, a real structural advantage over platforms where the entire application process is unpaid. micro1's process moves fast at the top, next-step emails typically arrive within 48-72 hours, though converting that into an actual paid assignment isn't guaranteed for every applicant.

Realistic expectations worth setting before you apply to either: data labeling work is project-based, not a steady job, pay varies by task type more than by platform, and a specific listing you see today may close or get replaced by the time you check back. Applying to both platforms in parallel, rather than waiting on one, is the practical way to avoid losing weeks to a single project ending or a single application going nowhere.

Other Platforms That Offer Data Labeling Work

Mercor and micro1 aren't the only places to find this work, and depending on your situation, one of these might fit better.

Each review linked above covers real pay data and the specific risks reviewers report for that platform, not marketing copy. For most readers starting from zero, Mercor's paid onboarding and micro1's current robotics labeling push are the strongest entry points covered on this page right now, but applying to more than one platform is still the practical move.

Worth Trying Instead

Frequently Asked Questions

Related Articles