Skip to content

Engineering · Salary Guide

Machine Learning Engineer Salary Guide (2026)

ML engineer compensation in the US is one of the most variable in tech — entry levels start around $115k and LLM research engineers at top AI labs clear $1M total comp. This guide breaks down the numbers by experience, city, and company — and the research vs. applied-ML split that decides where you land.

US national median (base salary)

$195k

Total compensation typically runs 30–60% above base at public tech companies.

Salary by experience level

Experience levelBase salary (US national)
Entry (0–2 yrs)$115k – $170k
Mid (3–5 yrs)$160k – $235k
Senior (6–9 yrs)$215k – $340k
Staff / Principal (10+ yrs)$310k – $600k

Salary by city

Base salary ranges adjusted by local market multipliers. Click through for the full city guide with tax notes and cost-of-living context.

CityAdjusted rangeMarket note
San Francisco$216k – $459kHighest base salaries in the US but COL roughly 2x national average.Full guide
New York$205k – $435kFinance premium for quant/fintech roles. Strong Big Tech satellite offices.Full guide
Seattle$192k – $408kNo state income tax. Dominated by Amazon and Microsoft; AWS roles pay a premium.Full guide
Austin$168k – $357kLower base salaries but strong take-home due to zero state tax and ~40% lower housing costs than SF.Full guide
Los Angeles$184k – $391kStrong media/entertainment tech market. Base salaries below SF but above most US markets.Full guide
Remote (US)$152k – $323kRemote-first companies typically pay at 85–105% of SF rates regardless of employee location.Full guide

Top employers and their ranges

OpenAI

PPUs replace RSUs; research-engineer roles top the market

$350k – $1.1M+ total comp
Google DeepMind (L3–L6)

Research engineer ladder pays above product SWE

$225k – $680k total comp
Meta AI (E4–E7)

GenAI org has standing ~15% uplift over general SWE

$240k – $900k total comp
Airbnb (L4–L6, ML)

Applied ML roles (pricing, search) sit just below research labs

$220k – $520k total comp
Uber (L5a–L7)

Strong applied-ML ladder; Michelangelo platform heritage

$230k – $560k total comp
Netflix (Senior ML)

Single-tier senior band; cash-heavy like other Netflix roles

$500k – $780k total comp

Skills that boost salary

LLM fine-tuning / RLHF experience+20% to +45%
Distributed training (Ray, DeepSpeed, Megatron)+15% to +28%
GPU kernels / CUDA / Triton+18% to +35%
Production ML platform ownership+10% to +20%
Evaluation and safety / red-teaming+12% to +22%

Negotiation tactics

  1. 1Separate research-engineer from applied-ML offers in your mind. Research-engineer offers at frontier labs (OpenAI, Anthropic, DeepMind) clear $500k–$1M+ but have tiny hiring bars; applied ML at large companies is easier to get but caps around $450k TC.
  2. 2Get competing offers from at least one frontier lab if at all possible. Even a non-accepted offer from OpenAI or Anthropic moves any other ML offer by 20–30%.
  3. 3Negotiate compute access, not just salary. At senior levels, a firm commitment of X GPU-hours/quarter for research time is worth more than a $20k base bump for your next career move.
  4. 4Publish. First-author NeurIPS/ICML/ICLR papers add 15–30% to offers at labs and Big Tech research orgs. One good paper can outperform two years of promotion velocity.
  5. 5Benchmark against MLE, not SWE, ladders. ML roles at most Big Tech companies sit on a separate, higher band than general SWE — confirm before accepting a 'leveled-in' offer that assumes parity.

FAQ

What is the average ML engineer salary in the US?

As of 2026, the median base salary for a US ML engineer is about $195,000, with mid-level ranges of $160k–$235k. Total compensation varies widely — applied ML engineers at Big Tech clear $300k–$450k TC, while research engineers at frontier AI labs reach $700k–$1.1M+.

How much do research engineers at OpenAI or Anthropic make?

Research engineers at OpenAI and Anthropic typically earn $350k–$700k total comp at the mid level and $700k–$1.1M+ at the senior level. Compensation is paid in base plus PPUs (OpenAI) or equity (Anthropic). Offers above $1M generally require a published research track record.

Do ML engineers get paid more than software engineers?

Yes. ML engineers earn roughly 15–25% more than same-level SWEs at most Big Tech companies in 2026, and the gap widens at research labs. Inside a single company, MLE and SWE may sit on separate ladders with different compensation bands.

Is it worth specializing in LLMs for salary growth?

In 2026, yes. LLM fine-tuning, RLHF, and distributed-training skills command the highest premiums in the market — 20–45% above general ML roles. The tradeoff is that the field moves fast; engineers who don't stay current on post-training methods see premiums fade within 12–18 months.

Sources: US Bureau of Labor Statistics — Occupational Employment and Wages, May 2025 · Levels.fyi aggregated offer data (Q1 2026) · H1B Salary Database (2025–2026 filings) · Built In salary reports (2026)

Land a higher-paying Machine Learning Engineer role

The resume is the first leverage point in a negotiation. Rolevanta tailors your resume to each job description and signals the seniority your numbers support.

Build Your Resume Free