How much do AI training jobs pay per hour?

Most roles on OpenTrain list an hourly rate in US dollars. The index shows the current median and typical range across all open roles, then breaks it down by role type and country. Specialist review in law, finance, and medicine sits at the top of the range. Audio and data collection roles sit near the bottom.

Where do these pay rates come from?

Every figure is calculated from the pay listed on open roles on OpenTrain at the last refresh. Nothing is estimated or self-reported, and the index refreshes every hour as roles open and close.

Does the rate shown mean I will earn that amount?

No. The listing you apply to is the source of truth. Rates on this page describe what open roles currently list. Actual earnings depend on the role, your qualifications, and the hours you work.

Why do entry-level roles sometimes pay more than intermediate roles?

Entry level describes the AI training experience required, not the difficulty of the work. Many entry-level roles are specialist roles that need a degree or professional background rather than prior AI training experience, and those roles list higher rates.

Are rates different by country?

Yes. Roles limited to specific countries often list rates aligned to that market. The country table shows the median for roles open to each country. Roles open worldwide are included in the overall figures.

AI training jobs Pay rates

AI training pay rates

What AI training work pays right now, calculated from every open role on OpenTrain and refreshed every hour.

Updated November 2, 2026 from 876 open roles. All rates in US dollars.

Median hourly rate: $65/hr; across all hourly roles
Typical range: $30/hr to $100/hr; 25th to 75th percentile
Paid hourly: 78%; of 876 open roles
Entry level: 64%; no prior AI training experience required

Hourly pay by role type

Sorted by median rate. Groups with fewer than 5 hourly roles are omitted.

Role type	Open roles	Median	Typical range	Entry level
Writing & Editing	6	$350/hr	$350/hr to $350/hr	100%
Legal & Finance	90	$120/hr	$90/hr to $180/hr	57%
Medical & Health	52	$90/hr	$70/hr to $115/hr	52%
Coding & Software	153	$68/hr	$50/hr to $100/hr	54%
Generative AI & RLHF	369	$60/hr	$30/hr to $90/hr	68%
Image & Video Annotation	32	$60/hr	$10/hr to $80/hr	72%
General Annotation	31	$60/hr	$60/hr to $94/hr	55%
Audio & Speech	74	$50/hr	$24/hr to $65/hr	91%
Translation & Localization	38	$30/hr	$25/hr to $39/hr	29%
Data Collection	30	$20/hr	$15/hr to $30/hr	77%

Hourly pay by country

Countries with the most open roles. A role open to several countries counts toward each of them.

Country	Open roles	Median	Typical range	Entry level
United States	501	$80/hr	$60/hr to $110/hr	59%
India	400	$70/hr	$36/hr to $100/hr	65%
Bangladesh	354	$70/hr	$55/hr to $100/hr	66%
Pakistan	350	$70/hr	$55/hr to $100/hr	67%
Brazil	318	$80/hr	$60/hr to $105/hr	63%
Indonesia	311	$78/hr	$60/hr to $105/hr	62%
Malaysia	309	$80/hr	$60/hr to $105/hr	62%
Thailand	308	$80/hr	$60/hr to $106/hr	62%
Germany	306	$80/hr	$60/hr to $110/hr	62%
Philippines	306	$80/hr	$60/hr to $105/hr	62%
Vietnam	304	$80/hr	$60/hr to $110/hr	62%
Sri Lanka	300	$80/hr	$60/hr to $110/hr	63%
Singapore	291	$80/hr	$60/hr to $110/hr	63%
Mexico	272	$80/hr	$60/hr to $115/hr	71%
Canada	266	$88/hr	$60/hr to $120/hr	70%
Egypt	263	$80/hr	$60/hr to $116/hr	71%
Nigeria	261	$80/hr	$60/hr to $119/hr	71%
Ghana	260	$80/hr	$60/hr to $119/hr	71%
Kenya	259	$80/hr	$60/hr to $119/hr	71%
Spain	256	$80/hr	$60/hr to $115/hr	70%

Rates are as listed on open roles hired through OpenTrain at the time of the last refresh. They describe current listings, not a guarantee of earnings. The listing you apply to is always the source of truth.

How to read this index

Median is the middle rate across roles that list an hourly rate. The typical range runs from the 25th to the 75th percentile, so half of all roles list a rate inside it. Entry-level share is the percentage of roles in a group that require no prior AI training experience.

What moves the rate

Three things explain most of the spread: the expertise a role requires, the language it needs, and where the client wants the work done. Specialist review in law, finance, medicine, and engineering pays the most. Coding and model evaluation sit in the middle. Transcription, data collection, and translation roles list the lowest rates but are also the most open to applicants without experience.

Apply through OpenTrain

Every role behind these figures is hired through OpenTrain. Create a free account, complete your profile once, and apply to any role that fits. Successful candidates may work on an OpenTrain project or be placed or referred to the client whose project they support.

Frequently asked questions

How much do AI training jobs pay per hour?: Most roles on OpenTrain list an hourly rate in US dollars. The index shows the current median and typical range across all open roles, then breaks it down by role type and country. Specialist review in law, finance, and medicine sits at the top of the range. Audio and data collection roles sit near the bottom.
Where do these pay rates come from?: Every figure is calculated from the pay listed on open roles on OpenTrain at the last refresh. Nothing is estimated or self-reported, and the index refreshes every hour as roles open and close.
Does the rate shown mean I will earn that amount?: No. The listing you apply to is the source of truth. Rates on this page describe what open roles currently list. Actual earnings depend on the role, your qualifications, and the hours you work.
Why do entry-level roles sometimes pay more than intermediate roles?: Entry level describes the AI training experience required, not the difficulty of the work. Many entry-level roles are specialist roles that need a degree or professional background rather than prior AI training experience, and those roles list higher rates.
Are rates different by country?: Yes. Roles limited to specific countries often list rates aligned to that market. The country table shows the median for roles open to each country. Roles open worldwide are included in the overall figures.