Where to find AI training work and paid research studies: platforms compared, and how to stop losing hours searching
AI training work is found on DataAnnotation, Outlier, Alignerr and Mercor; paid research studies on Respondent, User Interviews, Prolific, uTest, TestingTime and Userlytics. Each platform shows only its own tasks, so anyone who wants to earn regularly ends up checking five to ten sites a day. LazyBounty gathers open studies into one catalogue: on 26 September 2026 it lists 812 open tasks, 305 of them added in the last week.
What is open in the catalogue right now
Our own data, as of 27 September 2026
- 795open studies
- 284of them appeared this week
- 121countries among the open studies
- $40median payout
- $74average payout
- 314studies pay $50 or more
Most available in: United States — 199, United Kingdom — 49, Germany — 27, Israel — 24, Spain — 20
These numbers are counted from our catalogue at the moment you opened this page. The platform running the study pays you; LazyBounty never holds the money and never screens you.
Two kinds of platforms
AI training platforms hire people onto projects: you pass a test and then receive tasks for as long as the project runs. Paid research platforms work differently: each study is a separate event with its own screener survey, and you apply to each one individually.
So the search is different too. On an AI training platform, what matters is getting through the entry once. On research platforms, what matters is seeing new open studies regularly, because places fill fast and screeners close as soon as the researcher has enough people.
AI training platforms
Rates and requirements are taken from the platforms’ official pages on 26 September 2026. Each one is covered in detail in our article on AI training pay.
| Platform | Pay, in their words | Entry | Payout |
|---|---|---|---|
| DataAnnotation | from $20–30 an hour, specialists from $50–100 | bachelor’s or equivalent experience, test | PayPal |
| Outlier | rate shown before a project starts | ID, resume, LinkedIn, test | PayPal, Airtm, ACH weekly |
| Alignerr | average $80 an hour | experts, researchers, PhDs | weekly |
| Mercor | average contracted rate $109 | physicians, lawyers, engineers | per contract |
Paid research platforms
The "Our status" column shows our real relationship with each platform. "Partner" means a signed agreement. "In talks" means we are discussing an official connection with the platform and have had a reply. "Contacted" means we have written about a connection but have no substantive reply yet.
| Platform | Format | Pay | Our status |
|---|---|---|---|
| Respondent | interviews for consumers and professionals | most sessions $50–250, senior B2B roles $400+ | studies listed in our catalogue |
| uTest | bug hunting, test cases | depends on project and rating | projects listed in our catalogue |
| Prolific | online surveys and experiments | minimum £6/$8 an hour | separate, no agreement |
| User Interviews | interviews, focus groups, diaries | average over $45 | contacted |
| TestingTime | tests, interviews, focus groups | amount shown before you take part | in talks |
| Userlytics | recorded website and app tests | not published | in talks |
| Ipsos iSay | survey panel | points per survey | official partner since 22 Sep 2026; we are rewarded when a referred person signs up and completes a profile |
Which study formats pay more
Pay depends on format and length. User Interviews publishes ranges for its studies, and they show the overall picture well: in-person studies and multi-day diaries pay the most, short surveys are the easiest to get.
| Format | Length | Pay |
|---|---|---|
| surveys and testing | from 10 minutes to a few hours | $20–300+ |
| one-on-one interview | 30 minutes – 2 hours | $20–300+ |
| online focus group | 1–3 hours | $40–200+ |
| diary study | from 3 days to several weeks | $55–400+ |
| in-person study | 30 minutes – 2 hours | $40–1,000+ |
How to choose a platform: four questions
There are many platforms, and signing up for all of them at once makes little sense: each one needs its own profile, and a carelessly filled profile is worse than none, because screeners match on it and a wrong answer can quietly exclude you from studies you would have been perfect for. Before you register, answer four questions for yourself.
| Question | Where to find the answer | Why it matters |
|---|---|---|
| Does the platform work in your country | the countries page in its help centre | otherwise the account fails verification or gets no tasks |
| How does it pay out | the payouts section: PayPal, Payoneer, bank, gift cards | not every method is available in every country |
| Who is it looking for | the participant description and entry requirements | a test in the wrong field almost always fails |
| How long until the first payout | the payment timing section | anywhere from a few days to several weeks |
How to check that a platform is real
Real platforms pay you; you do not pay them. DataAnnotation states plainly that it will never ask workers for money and charges no signup fees, and the same rule holds for every honest platform. If "work" is offered in a first message on a messenger app, if you are asked to top up a balance to receive a payout, or if earnings are promised with no checks at all, it is a scam.
- the platform has a help centre with rules, timelines and payout methods
- the payment for a study is shown before you take part, not "after a deposit"
- you hand over documents only through the platform’s secure verification form
- you can find reviews of its payouts by searching the platform’s name
How much time the search takes
The main hidden cost in this work is not the study itself but the search. To avoid missing a well-paid interview you have to visit several platforms every day, open new studies, read the requirements and discard the ones that do not fit your country, language or profile. At half an hour a day, that is 15 unpaid hours a month.
That is why we calculate the rate honestly. The median stated rate in our catalogue is about $87 per session hour, based on 692 open Respondent studies on 10 September 2026. But that is the rate for an hour of conversation, not an hour of searching, and rejections are not included. Cutting search time is the simplest way to bring your real earnings closer to that number.
What LazyBounty does
LazyBounty gathers open paid studies into one feed and shows the ones that may fit your country and profile; the final selection is made by the platform. On 26 September 2026 the catalogue holds 812 open tasks: 664 Respondent studies and 148 uTest projects, 305 of them added in the last week. The feed also shows the panel of our partner Ipsos iSay in the countries where it operates.
We do not run studies ourselves and we do not pay for participation: you are paid by the platform that runs the study, under its rules. Our job is to help you see a suitable study sooner and not miss it.
- one feed instead of checking several sites
- filters by country, format and pay
- alerts about new matching studies
- your own “saved” and “applied” marks in one place
What LazyBounty does not replace
We are open about the limits. AI training platforms — DataAnnotation, Outlier, Alignerr, Mercor — hire onto projects through their own entry process, so their tasks are not in our feed: you need to sign up with them directly. We do not show Prolific either: its studies are visible only inside a participant account.
Paid studies with a screener survey, on the other hand, are exactly the format where an aggregator saves the most time: there are many of them, they close quickly and they are spread across different platforms.
How to combine it all
A practical setup that works for most people: pass the entry once on one AI training platform that fits your experience, and at the same time receive matching studies in one feed. AI training gives steady volume, research gives well-paid single sessions.
- strong written English — general AI training projects plus consumer interviews
- programmer — coding projects plus uTest testing and Respondent professional interviews
- doctor, lawyer, finance professional — specialist AI projects plus expert interviews where senior roles earn $400+
- no specialist background — profile-based studies and survey panels
Questions and answers
Where can I find AI training jobs?
On specialist platforms: DataAnnotation, Outlier, Alignerr, Mercor, and in the AI taskers group on Prolific. Each has its own entry test.
Where can I find paid research studies?
On Respondent, User Interviews, Prolific, uTest, TestingTime, Userlytics and others. LazyBounty gathers open studies into one feed: on 26 Sep 2026 it lists 812 open tasks.
Does LazyBounty pay for studies?
No. The platform that runs the study pays you, under its own rules. LazyBounty helps you find suitable studies faster.
Does LazyBounty list AI training tasks?
No. DataAnnotation, Outlier, Alignerr and Mercor hire through their own entry process, so you sign up with them directly.
Which platforms does LazyBounty officially work with?
Our official partner is Ipsos iSay, since 22 Sep 2026. We are in talks with TestingTime and Userlytics. Respondent studies and uTest projects are listed in the catalogue.
How often do new studies appear?
Constantly, but they also close quickly. In the week to 26 Sep 2026, 305 new tasks were added to the LazyBounty catalogue, so it is worth checking the feed daily or turning on alerts for new matching studies. Screeners often close within a day or two once enough people have applied.
Sources
- DataAnnotation: FAQs
- Outlier: FAQs
- Alignerr: Powered by Labelbox
- Prolific: Underpaying studies (minimum pay rate)
- User Interviews: Get paid for surveys, focus groups, and studies
- Respondent: What is the correct incentive amount for my research project?
- uTest: How can I earn money on uTest?
- TestingTime: FAQ Test users
- Userlytics: When will I get paid?
Useful on the site
Ready to try?
Fill in your profile once and see only the studies you actually qualify for.
Start the 7-day trial