How Much Do Machine Learning Engineer Jobs Pay in the USA? (September 2026)
As of September 2026, the median Machine Learning Engineer salary in the USA is $223,750 a year. Most roles pay between $158,400 and $303,950 (the 10th–90th percentile), drawn from 477 live salaried listings across 197 companies. All figures are converted to annual US dollars so different sources compare like-for-like.
How pay has moved
Median base salary by the month each role was posted, across current live listings. A monthly snapshot builds a firmer long-run trend over time.
Highest-paying employers
- Research Engineer, Machine Learning (RL Velocity)Research Engineer, Pretraining Scaling Research Engineer / Research Scientist, TokensML/Research Engineer, SafeguardsResearch Engineer, Life SciencesAll Anthropic roles →
- Machine Learning Engineer, IntegrityMachine Learning Engineer, Multimodal Perception and AuthenticationMachine Learning Engineer, API MulticloudMachine Learning Engineer, Distributed Data Systems - RoboticsAll OpenAI roles →
- Applied Machine Learning Engineer, EconomistMachine Learning Engineer, Frontier Data ProductsSoftware Engineer, Machine LearningAll Mercor roles →
- All xAI roles →
- All Kikoff roles →
Employers ranked by median advertised base among those with at least two salaried listings in the USA; each one’s best-paying open roles are listed beneath it.
Methodology: Machine Learning Engineer listings in the USA with a published salary, updated daily. Figures come from roles that are open right now. Ranges and single figures are converted to an annual US-dollar midpoint; hourly rates are annualised and out-of-band values dropped. Salary data is self-reported by employers and covers the roles that disclose pay — treat it as a directional guide, not a guarantee.