Mindrift vs RWS TrainAI: pay and jobs compared
Based on the 153 open roles we track on the two platforms, updated hourly. Figures are what each platform advertises.
- Mindrift advertises higher pay: a median of $35/hr, against $13/hr on RWS TrainAI (about 2.6 times as much). Most of Mindrift’s postings give only an “up to” rate; we count 70% of it, but actual pay may differ.
- RWS TrainAI has more open roles right now: 140 against 13.
- Different kinds of work: Mindrift’s roles are mostly AI Evaluation (15%); RWS TrainAI’s are mostly Multilingual (45%).
- Mindrift is more open worldwide: 54% of its roles accept applicants from any country, against 0% on RWS TrainAI (mostly Brazil, India).
| Open roles | 13 | 140 |
| Median advertised pay | $35/hr (“up to” rates) | $13/hr |
| Typical range (middle half) | $21–$61/hr | $8–$20/hr |
| Highest advertised rate | $100/hr | $36/hr |
| Roles that list pay | 69% | 99% |
| Paid per task | — | — |
| Main kinds of work | AI Evaluation 2 Coding 2 Data Annotation 2 | Multilingual 63 Transcription & Audio 48 Domain Experts 23 |
| Where they hire | Worldwide (54%) Worldwide 54% · named countries 8% · not stated 38% | Mostly Brazil, India (98%) Worldwide 0% · named countries 98% · not stated 2% |
| Remote | 100% | 100% |
Legitimate platforms never charge you to join, so there’s no cost to applying to both. How to spot scams.
Top roles on each
Mindrift
RWS TrainAI
Which should you choose?
Pick the platform whose work matches your skills. A specialist with a degree in medicine, law or a science will usually earn most on the platform hiring for that field, whatever the median says. If you can do the work on both, there’s no downside to applying to both and taking whichever offers projects first: work on AI training platforms comes in bursts, and having two sources keeps you busy.
Pay here is what each platform advertises in its postings. Real earnings depend on how many hours of work you’re given, and some roles pay per task rather than per hour. See how much AI training pays for more.