AI review specialties and rates

Specialties

All ten specialties.

How every specialty works

Every specialty runs on the same track. Practise free, as many times as you like. When you're ready, take that specialty's Cognition Test: 80 questions, pass mark 78%, one attempt per payment. Passing qualifies you for paid AI review work at the tier rate shown on the card, with your first tasks in your dashboard within 24 hours.

Not sure where you fit? Start with the Generalist practice test

Each specialty has its own practice test and its own Cognition Test. Pick the one where your strongest knowledge sits.

● Live, paid tasks available● Available to qualify - no current tasks

Basic tier

e.g. judge the ambiguous case: which of two answers serves the question better?

Generalist

Review AI outputs across everyday subjects: current affairs, basic geography, consumer reasoning, and the plausible-but-false claims models often produce. The entry route into Jobpeak, and the specialty where honest calibration counts for more than deep expertise.

Remote

Up to $30/hr

Start Practice Test

STEM tier

e.g. verify the proof, not just the answer.

Mathematics

Verify proofs, check working, and catch the errors that look right but aren't in AI-generated mathematics. Models often get the final answer right on fabricated working. The job is tracing the reasoning across algebra, calculus, statistics, and probability, and finding where it breaks.

Remote

Up to $40/hr

Start Practice Test

Physics

Review AI-generated physics work across mechanics, electromagnetism, thermodynamics, optics, and modern physics. The most common model failure is the right formula in the wrong regime, so the work rewards judgement on units, signs, and boundary conditions over rote recall.

Remote

Up to $40/hr

Start Practice Test

Chemistry

Review AI-generated chemistry across organic, inorganic, and physical chemistry: reactions, mechanisms, and stoichiometry. Models draw mechanisms that look plausible while breaking conservation rules; the job is reading what's drawn and judging whether the chemistry actually works.

Remote

Up to $40/hr

Start Practice Test

Biology

Review AI-generated biology across cell biology, genetics, ecology, evolution, and human physiology. The hardest model errors are scale errors, real mechanisms applied at the wrong level, so the work rewards keeping cellular and organismal reasoning straight.

Remote

Up to $40/hr

Start Practice Test

Coding tier

e.g. spot the bug a model just shipped.

Coding

Code review, debugging, and evaluation of AI-generated code across Python, JavaScript, TypeScript, SQL, and the other everyday languages. Models produce code that compiles but fails at runtime, misses edge cases, or calls APIs that don't exist. The job is catching it at speed, at the highest live rate ceiling on the platform.

Remote

Up to $100/hr

Start Practice Test

Professional tier

e.g. catch the confident error in a financial calculation.

Finance

Review AI-generated finance work at the Professional tier: valuation, financial statement analysis, and capital markets. The hardest errors are internally consistent models built on the wrong assumptions, whether the comparable, the discount rate, or the growth input.

Remote

Up to $50/hr

Start Practice Test

Accounting

Review AI-generated accounting work at the Professional tier: GAAP and IFRS interpretation, statement analysis, and audit logic. Models produce journal entries that pass surface inspection while misclassifying the underlying transaction; the work is reading past form to substance.

Remote

Up to $50/hr

Start Practice Test