Expert gig work training and evaluating AI models, aggregated from the marketplaces that actually pay for it.
1,347 open listings across 6 marketplaces, refreshed daily.
Filtering by rate hides per-task listings, which have no hourly equivalent.
426 listings
Clear filters
Turkish Translator Overview Automated metrics can't judge whether Turkish text actually sounds right: tone, idiom, cultural nuance, and the thousand small choices a native speaker…
Spanish Translator Overview Automated metrics can't judge whether Spanish text actually sounds right: tone, idiom, cultural nuance, and the thousand small choices a native speaker…
Software Engineer Overview The newest models can vibe-code a whole feature in minutes. Someone has to check whether that code actually works, and that someone is you. You'll run ag…
Quantitative Analyst Overview Models handle textbook finance fine and fall apart on the real thing: mispriced risk, subtly wrong derivations, and models that ignore market microstr…
Portuguese Translator Overview Automated metrics can't judge whether Portuguese text actually sounds right: tone, idiom, cultural nuance, and the thousand small choices a native sp…
Physicist Overview Models can recite physics and still get it wrong: dimensional errors, misapplied approximations, and derivations that skip the step where the physics actually ha…
Physician Overview Medical questions are where model mistakes matter most: outdated guidance, missed contraindications, and confident answers that no clinician would sign off on. A…
Mathematician Overview Frontier models still fail at rigorous multi-step mathematics: invalid derivations, unjustified leaps, and confidently wrong final answers. Automated checks…
Machine Learning Engineer Overview Models are surprisingly bad at reasoning about themselves: training dynamics, evaluation design, data pipelines, and deployment trade-offs. Plaus…
Indonesian Translator Overview Automated metrics can't judge whether Indonesian text actually sounds right: tone, idiom, cultural nuance, and the thousand small choices a native sp…
Hindi Translator Overview Automated metrics can't judge whether Hindi text actually sounds right: tone, idiom, cultural nuance, and the thousand small choices a native speaker make…
Healthcare Operations Specialist Overview Models struggle with the practical side of healthcare: billing codes, insurance workflows, scheduling, and the administrative details that…
Generalist Overview Most of what models get wrong isn't specialist knowledge. It's careful reading, clear writing, and common-sense judgment, and that's exactly what this role supp…
Data Scientist Overview The newest models can spin up a full analysis in minutes — load the data, pick a method, produce charts and a confident conclusion. Someone has to check whe…
Chemist Overview Models routinely propose impossible mechanisms, unsafe procedures, and syntheses that no bench chemist would attempt. As a Chemist you'll evaluate AI chemical reas…
Biologist Overview Models blur the line between what's established, what's contested, and what's simply made up in the life sciences, and they cite papers that don't exist. As a Bi…
Attorney Overview Models produce fluent legal analysis that is often subtly wrong: invented citations, misapplied precedent, and jurisdictional confusion. A plausible wrong answer…
Arabic Translator Overview Automated metrics can't judge whether Arabic text actually sounds right: tone, idiom, cultural nuance, and the thousand small choices a native speaker ma…
Accountant Overview Models produce confident accounting answers that fail on the details: misapplied standards, broken reconciliations, and journal entries that don't balance. As a…
Short-Form Video Creators: Get Paid Without Pitching for It. Use your video editing skills to train AI models. Work remotely, set your own hours, and earn up to $15/hr with Outlier…
Statistics Experts Wanted to Train the Next Generation of AI Use your statistics expertise to train AI models. Work remotely, set your own hours, and earn up to $150/hr. Statistics…
Quant Finance Experts: Earn Up to $150/hr Helping AI Understand Markets Put your quantitative finance expertise to work. Train AI models, earn $40–$150/hr, fully remote, no minimum…
Personal Finance Experts Wanted to Train Advanced AI Earn up to $90/hr applying your CFP expertise to train AI on real client financial planning scenarios. Outlier AI needs seasone…
Ophthalmologists: Help Define How AI Sees Medicine What is Outlier? A platform for building AI with expert human input. Discover Outlier AI, join a community for freelancers, and s…
AI labs need domain experts to write, review and grade the material their models learn from. A radiologist checks a diagnostic explanation, a tax accountant writes the reasoning behind a filing, a backend engineer builds a test environment an agent has to solve. That work is bought through a handful of specialist marketplaces, it is almost always remote, and it is almost always paid by the hour or by the task.
The problem is that the work is scattered. Each marketplace runs its own board, its own titles and its own pay conventions, so finding the roles that fit you means checking six sites and mentally converting between them. This page collects the listings into one feed with one set of filters, refreshed daily, and links each one straight through to the marketplace that posted it.
Search by skill, filter by pay, commitment or experience level, and see every marketplace side by side instead of one board at a time.
Every listing links to the original posting. You apply, onboard and get paid by them directly. Skillora does not sit between you and your earnings.
Most of these marketplaces screen with a recorded or live interview. Run the same interview with Skillora first and get feedback on your answers.
Write and review code, build agent environments, grade model output.
Evaluation design, dataset curation and reinforcement-learning work.
Clinical reasoning, diagnostic review and medical exam authoring.
Contract analysis, case reasoning and jurisdiction-specific review.
Modelling, valuation walkthroughs and accounting treatment review.
Translation, transcription, speech recording and native-fluency review.
Physics, chemistry and biology problem authoring and verification.
Consulting cases, operations workflows and process documentation.
Every listing links back to the marketplace that published it. You apply, onboard and get paid by them directly.
Expert contract work across medicine, law, finance and engineering, mostly hourly and often at the top of the rate range.
See Mercor listings
AI-vetted talent network placing engineers and domain specialists on model-training projects.
See micro1 listings
Large catalogue of AI training roles, heavily weighted towards software engineering and language work.
See Alignerr listingsDomain verticals with published rates, covering coding, maths, writing and specialist subjects.
See Outlier listingsTask-based annotation and model-response review, with an accessible entry path for new contributors.
See DataAnnotation listings
Paid expert studies and evaluation projects, priced per approved study rather than hourly.
See Terac listingsIt is paid work producing or evaluating the data that AI models learn from. In practice that means writing expert answers in your own field, grading model output against a rubric, building coding or tool environments for an agent to solve, recording speech, or reviewing another contributor’s work. It is contract or task-based rather than salaried, and the marketplaces that buy it recruit for specific domains rather than for general labelling.
Rates vary far more than in a normal job market because they track scarcity of expertise, not seniority. General contributor work sits in the $15 to $30 per hour range. Specialist domains such as medicine, law, quantitative finance and senior software engineering routinely post between $60 and $150 per hour, and some expert roles go higher. Some work is priced per task instead of hourly, in which case we show the task fee as published rather than inventing an hourly equivalent.
Usually not. These roles hire for depth in a subject, not for ML skills, because the whole point is to get expertise into the model that the model does not already have. A practising lawyer or a cardiologist is a stronger candidate for a domain role than a generalist who knows how transformers work. Software engineering roles are the exception, since the work itself is writing and reviewing code.
Almost always. The overwhelming majority of listings in this feed are fully remote and asynchronous, which is why they suit contractors, people between roles and anyone doing this alongside another job. Where a listing states a country restriction we show it on the card, and where it states none we say so plainly rather than claiming it is open worldwide.
Mercor, micro1, Alignerr, Outlier, DataAnnotation and Terac. These are the platforms that publish expert AI-training work openly. We refresh the feed once a day and keep listings that have closed accessible, marked as no longer listed, so a link you saved never dies silently.
No. Browsing and applying are free, and you are never asked for an account to see a listing. Where a marketplace runs a referral programme, Skillora may earn a fee if you sign up and work through our link. It costs you nothing, we say so on every listing where it applies, and it has no effect on how listings are ranked. The ranking code has no knowledge of which sources pay.
Every marketplace screens before assigning work, most of them with an interview that is recorded or run by an AI interviewer, followed by a paid trial task. The screen is the part most candidates underestimate: it is where you have to explain your reasoning out loud in your own domain. Practising that specific format is the single highest-leverage thing you can do before applying.
The screen is usually a recorded or live interview in your own domain, followed by a paid trial task. Run that interview with Skillora before you take the real one and get feedback on how you explained your reasoning, not just on what you said.
Practise the screen free