Remote data annotation jobs
Data annotation means labeling images, video, text and audio, or rating outputs using a defined rubric. It creates structured data for AI training. Broader AI training roles also involve writing prompts, reference answers and specialist evaluation tasks.
Open roles right now
All jobsNo matching roles are open at this moment — browse all openings.
What the work involves
Most annotation tasks fall into four buckets: drawing boxes or masks around objects in images and video, transcribing and tagging audio, labeling text, and rating outputs against a defined rubric. You get written guidelines per project, and your work is spot-checked for quality. Accuracy matters more than speed.
Who hires annotators
Frontier labs rarely hire annotators directly. They contract through platforms like micro1, Mercor, Ethos and Turing. One application to a general pool usually beats applying to a dozen single roles.
Before you apply
Read what data annotation is and how to get your first job. Nearly every platform starts with an unpaid assessment — our assessment tips cover the mistakes that fail most applicants.
Looking for higher pay? AI training jobs and RLHF roles pay more for domain expertise.