Monthly report · November 2026
State of AI training pay
What 2,180 live roles on Mercor, micro1, Handshake AI, Alignerr, Appen and OpenTrain AI paid, by field, platform, credential and language. Data as of November 1, 2026.
By Michael Cortez, founder and editor
- Median hourly
- $60
- Top quarter from
- $90
- Roles at $100+/hr
- 20.5%
- Highest listed
- $450
Key findings
- 01
The median listed rate fell $2.50 an hour since October
On the 3 platforms tracked the same way in both months, the median listed rate was $65 in October 2026 (restated to this edition's method) and $62.50 now, a 4% drop. This edition adds OpenTrain and Appen (702 roles at a $52.50 median; 51 roles at a $35 median), so the all-platform median of $60 isn't comparable with October's. micro1 is left out: October covered 24 hand-picked micro1 roles; from October 30 we read micro1's full referral board (333 roles).
- 02
The typical role pays $60 an hour
Half of the 1,813 roles with an hourly rate pay more than $60 and half pay less. 20.5% pay $100 an hour or more, and 12.7% pay under $25.
- 03
12% of roles state only a maximum rate
264 of 2,180 live roles (all from Handshake AI) list pay as "up to" a figure, with a median ceiling of $80 an hour. A ceiling says nothing about the typical rate, so these roles are counted but kept out of every median here.
- 04
Pay varies far more by field than by platform
Law roles have the highest median at $115 an hour (138 roles); voice & audio roles have the lowest at $26 (240 roles), a 4.4x spread. Among platforms with 50+ hourly roles, medians range only from $52.50 to $80.
- 05
Roles that require a medical license have a median of $150 an hour
Roles that require an MD or physician license have a median of $150 an hour (64 roles), 2.5x the overall median. Against the same median: PhD roles $82.50 (+38%, 114 roles) and master's degree roles $70 (+17%, 111 roles).
- 06
Mercor has the highest median among platforms with 50+ hourly roles
OpenTrain posts the most roles (702) at a $52.50 median. Among platforms with 50+ hourly roles, Mercor has the highest median, $80, across 282 roles.
- 07
Native-speaker roles pay 50% of the typical rate
Roles that ask for a native speaker have a median of $30 an hour (73 roles), against $60 overall. Among languages with 10+ hourly roles, Spanish pays most ($50, 46 roles) and Hindi, Portuguese and French least ($22.50 each).
What changed since October
Like for like: only the 3 platforms covered the same way in both editions, so a newly added platform can't pass for a change in pay.
micro1 is left out: October covered 24 hand-picked micro1 roles; from October 30 we read micro1's full referral board (333 roles). The jump is a change in coverage, not in pay, so micro1 is left out of the comparison on both sides.
- Median, same platforms
- $65 → $62.50
- Roles, same platforms
- 1,070 → 1,094
- New this month
- OpenTrain, Appen
By platform
| Platform | October | Now | Change |
|---|---|---|---|
| Mercor | $80 | $80 | same |
| micro1 | $90 | $75 | coverage changed |
| Alignerr | $60 | $60 | same |
| OpenTrain | – | $52.50 | new |
| Appen | – | $35 | new |
| Handshake AI | – | – | new |
Biggest moves by field
- Medicine & health$60 → $90 (+50%)
- Languages & translation$80 → $50 (−37%)
- Design & media$55 → $70 (+27%)
- Law$85 → $95 (+12%)
- Physics$70 → $60 (−14%)
- Business & operations$60 → $57.50 (−4%)
Same platforms only; fields with at least 10 hourly roles in both months.
How pay is spread
Hourly roles by the midpoint of their listed range, in $25 bands.
By credential
Median hourly by what the listing requires. Small number is the roles with an hourly rate.
- MD / licensed physician$150 64
- PhD$82.50 114
- Master's degree$70 111
- No stated credential$60 1421
- Native speaker$30 73
By platform
Line marks the overall median.
- Mercor$80 282
- micro1$75 322
- Alignerr$60 516
- OpenTrain$52.50 683
- Appen$35 10
Not charted: Handshake AI lists only ceilings (264 roles, median up to $80/hr), so it has no comparable median.
By field
Fields with at least five hourly roles. Line marks the overall median.
- Law$115 138
- Medicine & health$90 139
- Machine learning$90 61
- Math & statistics$82.50 45
- Chemistry$82.50 39
- Finance & accounting$80 118
- Mental health$80 7
- Physics$75 34
- Biology & life sciences$70 27
- Software & coding$65 188
- Humanities & policy$65 11
- Engineering$62.50 117
- Writing & editing$60 23
- Business & operations$60 124
- Cybersecurity$60 50
- Design & media$55 31
- Generalist$50 140
- General STEM$50 13
- Energy & earth sciences$50 18
- Data & analytics$40 71
- Teaching$30 7
- Languages & translation$30 172
- Voice & audio$26 240
By language
Languages with at least three hourly roles.
- Filipino$80 4
- Hebrew$80 4
- Swedish$60 4
- Norwegian$55 5
- Danish$55 3
- Spanish$50 46
- Finnish$50 4
- Turkish$47.50 6
- Japanese$40 18
- Arabic$37.50 11
- Polish$35 4
- Chinese$32.50 30
- Greek$32.50 4
- Vietnamese$30 9
- Indonesian$30 6
- Russian$30 5
- Hungarian$30 3
- Czech$30 3
- German$27.50 37
- Italian$27.50 14
- Thai$27.50 4
- Dutch$26 14
- Korean$26 19
- Catalan$26 4
- Malay$25 3
- Tamil$25 7
- Romanian$25 3
- Telugu$22.50 5
- Hindi$22.50 10
- Portuguese$22.50 22
- French$22.50 21
- Bengali$22.50 7
- Marathi$22.50 3
- Gujarati$20 4
- Urdu$16 5
- Punjabi$15 4
Best-paying roles this month
- 1OncologistMercor · Medicine & health$350-$450/hr
- 2RadiologistMercor · Medicine & health$350/hr
- 3CardiologistMercor · Medicine & health$350/hr
- 4AnesthesiologistMercor · Medicine & health$350/hr
- 5Orthopedic SurgeonMercor · Medicine & health$350/hr
- 6Luxury Automotive Strategy ExpertMercor · Generalist$300-$350/hr
- 7Breast Imaging RadiologistMercor · Medicine & health$300/hr
- 8Machine Learning Evaluation SpecialistAlignerr · Machine learning$200-$400/hr
- 9Domain Researcher, Finance (M&A)micro1 · Finance & accounting$192-$385/hr
- 10Domain Researcher, Legalmicro1 · Law$192-$385/hr
Roles listed as of November 1, 2026; many have since closed. See what's open now.
How we built this
- Every live listing on the public job boards of Mercor, micro1 (hand-captured), Handshake AI, Alignerr, Appen and OpenTrain AI as of November 1, 2026. Alignerr postings that differ only by location are counted once.
- Each role counts once, at the midpoint of its listed hourly range. Per-task and one-time pay are excluded: there's no honest hourly figure for them.
- 264 roles list only a ceiling (“up to $X/hr”, median up to $80). A ceiling has no midpoint, so they are counted in role totals but left out of every median.
- Listed rates are what platforms advertise, before tax and before any time limits. What people actually earn can differ.
- This edition's numbers were frozen on November 1, 2026 and won't change.
- Corrected November 1, 2026: The headline figures are unchanged. The top-paid roles list had included two Handshake AI "up to" ceilings; it now uses midpoints only, like every median here. October's field figures in the month-over-month comparison now use the field categories in use since October 28, so each field is compared under the same definition in both months, and a move is shown only when both months have at least 10 priced roles. The language finding now names all three languages tied for the lowest median.
Cite or reuse this report
Free to quote, chart and republish under CC BY 4.0. Please credit AI Training Pay and link to this page.
Writing about AI training work? Every figure comes from public job listings, frozen on November 1, 2026, and the charts above are ready to use. Questions: hello@aitrainingpay.com. Quotable findings, a fact sheet and who to interview are on the press page.
AI Training Pay, "State of AI training pay: November 2026," November 1, 2026, https://www.aitrainingpay.com/reports/2026-11




