How much does an AI Research Engineer make?

The median AI Research Engineer salary is $175,000 per year, from 236 job postings that disclosed pay across 84 companies in October 2026. Most fall between $129,000 and $215,000.

Which companies pay AI Research Engineers the most?

Of employers with at least three open AI Research Engineer postings, OpenAI posted the highest median range at $307,500.

Do remote AI Research Engineer jobs pay less?

In this sample remote-eligible postings had a median of $159,803 against $213,600 for onsite — 25.2% less.

What skills do AI Research Engineer jobs require?

Across 300 AI Research Engineer posting descriptions read in October 2026, the most frequently required were: Python, Software Engineering, LLM, Machine Learning, AWS. The top one appears in 45% of postings.

AI Research Engineer Skills and Job Description (October 2026)

Name: AI Research Engineer salary data, October 2026
Creator: FitToHire

Python is asked for by 45% of AI Research Engineer employers, the most common requirement across 300 live postings read in October 2026. Then Software Engineering, LLM, Machine Learning.

Median posted pay is $175,000, most postings between $129,000 and $215,000 — from 236 postings that disclosed it, across 84 companies.

Posted pay is up 3.4% since 2026-09-27 ($169,233 → $175,000).

Check your résumé against these →

These are ranges employers published in their own job postings — not salaries people typed into a form years ago.

What AI Research Engineer postings actually ask for

Ranked by how many of the 300 posting descriptions mention them.

Python 45%

Software Engineering 24%

LLM 20%

Machine Learning 18%

AWS 16%

PyTorch 15%

Version Control 15%

Electrical Engineering 14%

Java 14%

Azure 14%

Data Science 14%

SQL 13%

Docker 11%

CI/CD 11%

Kubernetes 11%

Distributed Systems 10%

Cloud Platforms 10%

AI Tools 10%

Go 9%

Data Models 9%

Data Pipelines 9%

Artificial Intelligence 9%

Data Analysis 8%

Infrastructure As Code 8%

Would your background clear the bar?

These are the requirements AI Research Engineer postings list. Paste one of those postings and your résumé — FitToHire checks you against every requirement and shows the evidence behind each verdict.

15 companies in this sample have AI Research Engineer postings open right now.

Check your fit →

Who is hiring AI Research Engineers

15 companies with open postings in this sample.

OpenAI

$307,500 3

Anthropic

$307,500 3

Shield AI

$290,000 3

Information Technology Senior Management Forum

$289,900 5

NVIDIA

$241,812 6

Deloitte

$241,800 43

What AI Research Engineers earn

10th $85K 25th $129K median $175K 75th $215K 90th $242K

$70K $288K

Each bar is a pay band. Height is how many of the 84 employers who published a salary fall in it, counted once each.

Pay bands in numbers

Employer median pay by band
Pay band	Employers
$70,000–$91,750	9
$91,750–$113,500	8
$113,500–$135,250	5
$135,250–$157,000	12
$157,000–$178,750	11
$178,750–$200,500	9
$200,500–$222,250	16
$222,250–$244,000	6
$244,000–$265,750	1
$265,750–$287,500	7
Total	84

AI Research Engineer vs adjacent roles

Roles people compare this one against, taken from what they search.

ai engineer
ai research scientist
ml engineer
ai developer
data engineer

Remote vs onsite pay

Median of remote-eligible postings against the rest.

Remote-eligible

$159,803

Onsite

$213,600

56 remote-eligible postings against 181 onsite. Onsite pays 25.2% more in this sample.

Pay by experience level

Level is taken from the posting title, so only postings that state one are counted.

Senior

$241,800 84

Lead / Manager

$216,300 43

Principal / Staff

$225,000 22

Pay by location

Metros with at least three salaried postings.

New York, NY

$220,249 20

San Francisco, CA

$225,000 19

Remote, US

$150,000 13

Menlo Park, CA

$220,498 9

$155,000 7

San Diego, CA

$232,500 7

Santa Clara, CA

$235,750 7

Atlanta, GA

$162,505 6

Mountain View, CA

$195,750 6

Cupertino, CA

$214,000 6

Los Angeles, CA

$160,500 5

Seattle, WA

$186,375 4

Where this data comes from

300 live postings analysed
236 (79%) published a salary range
71 (24%) were remote-eligible
123 distinct employers
Requirements read from 300 full posting descriptions
Each employer gets one vote in the median, so a company running a large hiring campaign cannot move it
Ranges are converted to annual figures; hourly and contract postings are normalised before comparison

How the requirements were read

Requirements are not keyword counts. Each posting is read by our extraction engine, which works out what the employer is actually asking for and separates a requirement from a section heading or a line of boilerplate. Two postings asking for the same thing in different words are counted once, because our skills graph knows which phrasings mean the same thing and which imply others.

The share beside each requirement is the percentage of distinct EMPLOYERS in that sample who asked for it, counted one vote per company so a firm posting the same role many times cannot inflate its own requirements. Measured, not estimated, and read from the posting text rather than counted as keywords.

Because every reading is dated, the movement figures above are real month-over-month change. Sites built on self-reported salaries cannot produce that number: their data has no posting date.