Browse open subjects
Coding Specialist
Review code examples, test logic, and help early-stage models output cleaner code.
Generalist Specialist
Check response tone, verify basic facts, and review general written text.
Law Specialist
Review summaries of legal documents, check contract logic, and verify statutory references.
Mathematics Specialist
Review step-by-step equations, check algebraic proofs, and find errors in mathematical logic.
Medicine Specialist
Review medical explanations, check clinical reasoning examples, and ensure health information is safe.
Physics Specialist
Review derivations, check electromagnetic calculations, and verify explanations of physical concepts.
Finance Specialist
Review corporate finance calculations, check market analyses, and verify mathematical logic.
Accounting Specialist
Review balance sheets, check GAAP/IFRS calculations, and verify general bookkeeping steps.
Bilingual Specialist
Review translation quality, check regional phrasing differences, and verify translated texts.
Chemistry Specialist
Review reaction steps, check chemical equations balancing, and verify molecular logic.
Biology Specialist
Review descriptions of cellular pathways, check genomic data, and verify scientific logic.
Real, thoughtful work.
Language models often make mistakes on complex rules and calculations. We focus on finding these errors rather than doing simple data entry.
Focus on accuracy
Getting the right answer is what matters most. You will check facts, review steps, and verify code.
Direct improvements
Your feedback directly helps update the models, teaching them how to solve problems step-by-step.
Work on your schedule
There are no daily meetings or set hours. Log in and work on projects whenever you have time.
Clear hourly pay
Your hours are tracked on your dashboard, and payouts are processed weekly.
How it works
1. Quick test
Take a short assessment to show your expertise in your chosen field.
2. Choose tasks
Once approved, you can pick any active project from your dashboard.
3. Get paid
Complete tasks on your own time, track your earnings, and get paid weekly.
What a task looks like
Here is an example of a typical workflow. You'll review a prompt, check the model's response, and write a correction if it made a mistake.
Active Directive
Draft highly specific, complex prompts in your field to test the boundaries of a frontier AI's capabilities. You'll construct scenarios that require multi-step reasoning, precise calculations, or structural syntax compliance.
PROMPT_COMMAND > Explain the mechanism of a nucleophilic acyl substitution under acidic vs basic conditions, citing stereochemical constraints.
Identify whether the model accurately outlines the protonation steps, recognizes tetrahedral intermediate stability, and identifies stereochemical preservation or inversion correctly.
System objective: Verify logical consistency under variable pH parameters.
Who we are looking for
Language models are good at basic writing, but they struggle with specialized topics. We want to work with scholars, students, and professionals.
Academics and students
Great for researchers, graduate students, and teachers in subjects like science, history, or math.
Developers and programmers
Ideal for software engineers and students who enjoy debugging, checking code efficiency, and analyzing logic.
Subject experts
Built for professionals like lawyers, accountants, financial analysts, and healthcare practitioners who want to consult on technical topics.
Linguists and translators
For bilingual writers and translators who can check translation accuracy, cultural nuances, and grammar.