Image Tagging & Annotation

Label images so machine-learning systems can learn what they show.

What the work is

Companies that train AI systems need huge volumes of labelled images — boxes drawn around cars, tags on clothing items, transcripts of street signs. As an image tagger you log into a labelling platform, get a queue of images, and apply the right tags or boxes according to a strict guideline document. Pay is modest but very steady, and the work is genuinely beginner-friendly. The number-one skill that gets you promoted to higher-paying tiers is reading the guideline document carefully and applying it consistently — most people skim it, which is why the small percentage who do not are paid more.

Day-to-day responsibilities

• Apply tags or bounding boxes per the guideline.
• Maintain a high accuracy rate to keep your tier.
• Flag ambiguous images for a reviewer instead of guessing.

Tools you actually need

• A laptop
• Steady internet
• Sharp eyesight

How beginners get hired

The most reliable hiring channels for this role today are listed below. Sign up on one or two — not all of them at once — and complete each profile fully before applying. Half-finished profiles convert poorly even when your skills are real.

• Remotasks
• Clickworker
• Appen
• Lionbridge

Beginner tips that actually work

• Re-read the guideline at the start of every shift — they update silently.
• Use a calibrated monitor or at least a non-tinted screen.
• Take a real five-minute break every hour. Eyes get tired and accuracy drops.

How and when you get paid

Weekly via PayPal or Payoneer on most platforms. See our guide to remote payment methods for how each option works in your country, and how to avoid the verification delays that catch most beginners off guard.

Country-specific guides

Pay norms, popular hiring platforms, and payment routes vary a lot by country. Pick yours below for the version of this guide tailored to where you live.