Two AI-training platforms side by side, from the roles they have open today and what we know about how each hires and pays. Numbers refresh daily.
| Alignerr | Mercor | |
|---|---|---|
| Open roles | 50 | 297 |
| Median top rate | — | $100/hr |
| Middle half of rates | — | $65/hr – $120/hr |
| Highest advertised | — | $450/hr |
| Open worldwide | 0% of roles | 65% of roles |
| Hires most for | Design & creative, Software engineering, Business & consulting | Science & research, Software engineering, Medicine & healthcare |
| Kinds of work | — | — |
| After you apply | Hiring is by tier: have evidence of your domain knowledge ready, since expert roles go to contributors who can verify it. | Expect an AI-led interview of 45–90 minutes, the longest of the major platforms. Most applicants are rejected; if you pass, work comes in intensive 2–8 week engagements. |
| How it pays | — | Weekly via Stripe |
Rates are what open listings advertise (median of the top of each range), not earnings. How we measure pay
Labelbox's expert network for RLHF and model-alignment work.
Frontier-lab SME placement at the highest pay tier in the market ($80–$200+/hr).