Role guides
AI job types, explained by the work and the pay
The title on a listing rarely tells you what the day looks like. These guides separate the seven roles that dominate AI training and evaluation work, with published pay bands and the screen each one uses.
AI Trainer jobs
$20–$100+/hr
AI trainers rank answers, write reference responses, and grade model output against a rubric. Pay runs from about $20/hr on generalist tracks to $100+/hr for credentialed experts.
AI Evaluator jobs
$22–$150/hr
AI evaluators judge model output against a rubric and defend the score in writing. It pays more than annotation because the justification is the product.
AI Tutor jobs
$15–$150/hr
AI tutoring is explanation work: writing step-by-step solutions, correcting model reasoning, and adapting an answer to the learner. Pay tracks the subject and the credential.
Data Annotator jobs
$15–$40/hr
Data annotators label images, text, audio, and video for training sets. It is the lowest-barrier entry into AI work and the clearest route to higher-paying evaluation tracks.
AI Code Reviewer jobs
$25–$200/hr
AI code reviewers judge generated code, write test cases, and explain failures. Pay runs from $25/hr on generalist tracks to $200/hr on specialist engineering networks.
AI Linguist jobs
$15–$40+/hr
AI linguists review translations, rate fluency, localize prompts, and explain grammar choices. Pay tracks the scarcity of the language pair more than the degree.
AI Domain Expert jobs
$50–$200/hr
Domain experts review model output in their profession. Medicine, law, finance, and advanced STEM pay $50–$200/hr because verification is expensive and a wrong answer carries liability.
How the tracks differ
The title on a listing is a weak signal. Two roles called AI trainer can pay $20/hr and $70/hr because one is generalist preference ranking and the other is specialist evaluation. Read the task description and the track, then compare the band. The track is the unit that platforms price.
Entry roles reward consistency: annotation, generalist ranking, language review. Specialist roles reward a defensible judgment: evaluation, code review, tutoring in a subject you can explain three levels deep. Expert roles reward verification: a licence, a degree, or a documented senior record. Each step up requires a harder assessment, not a longer resume.
Language and domain expertise are the two clearest shortcuts. A scarce language pair or a regulated credential narrows the applicant pool, and platforms pay for that scarcity. If you already hold one, apply to the expert track instead of competing in the generalist queue.
Not sure which role fits?
Start with your evidence. A licence, a degree, a code portfolio, or native fluency each opens a different track. The track matters more than the job title.
Compare the pay bands. The salary guide puts every track on one page, with the source for each band and the platform that publishes it.
Check the platform. Two platforms can pay the same headline rate and differ on payment timing, screening, and work stability. Compare before you apply.
Find the roles that match your track
Upload a resume and get every live role ranked by fit score, with published pay attached. Free, no account, results in under 20 seconds.