# Do remote jobs pay less? Yes, 13% less, and remote work isn't the reason

Published 2026-07-28. Updated 2026-07-28. [Hotfix Blog](https://hotfix.jobs/blog)

Remote engineering jobs advertise about 13% below onsite ones. Sort tech employers by what they pay and the reason shows up immediately: the lowest-paying quarter post half their roles remote, the highest-paying quarter post under a third. Inside any one company, remote and onsite pay the same to within a couple of percent.

Do remote jobs pay less in tech? Yes. Remote engineering jobs advertise about 13% below onsite ones at entry and mid level. But the discount is not attached to remote work, and a job seeker who reacts by dropping the remote filter is solving the wrong problem.

We sorted employers into four groups by their own median advertised engineering pay. The lowest-paying quarter posted 50.1% of their engineering roles as remote. The highest-paying quarter posted 29.4%. The gradient runs cleanly through all four groups, and it is the whole story.

Then we ran the comparison the other way: remote against onsite jobs at the same company, at the same seniority level. The gap there is 0.3%, and the true figure sits somewhere between a 2.6% discount and a 2.0% premium. Nothing close to 13% survives.

The remote filter is not pricing your commute. It is sorting you into a cheaper part of the employer market.

## The companies that pay least hire remote most

Chart: Share of engineering jobs posted as remote, with employers grouped into quarters by their own median advertised pay.

- Lowest 25% ($169k): 50.1%
- Second 25% ($197k): 35.9%
- Third 25% ($217k): 30.6%
- Highest 25% ($257k): 29.4%

Of the 213 companies with eight or more engineering jobs in this window, the cheapest quarter post remote roles at roughly 1.7 times the rate of the most expensive quarter. Narrow the same test to senior engineering only and it steepens: 54.6% remote at the bottom against 27.8% at the top.

This is what produces the headline gap. Remote jobs are drawn disproportionately from employers who pay less for every role they post, remote or not.

## What we measured

Hotfix ingests jobs directly from applicant tracking systems including Greenhouse, Lever, Ashby, and Rippling. This analysis covers US jobs in eight core engineering functions (backend, frontend, fullstack, DevOps, security, machine learning, data, and mobile) posted between January 1 and July 27, 2026, restricted to jobs carrying a complete advertised salary range above $40,000. Pay figures are the midpoint of the advertised range. Every comparison uses medians.

The within-company test pairs each company against itself: for a given employer at a given seniority level, we compare the average advertised midpoint of its remote jobs against its non-remote jobs, then look at the distribution of those gaps across 204 company-and-level pairings at 128 companies.

## The raw gap, which is what everyone else reports

Compare remote against onsite across the whole market and the penalty looks real and large:

| Level  | Remote   | Onsite   | Gap    |
| ------ | -------- | -------- | ------ |
| Entry  | $147,500 | $170,000 | -13.2% |
| Mid    | $180,000 | $207,500 | -13.3% |
| Senior | $191,950 | $205,000 | -6.4%  |
| Staff  | $237,263 | $242,500 | -2.2%  |

JobLeads published a study in 2026 running exactly this comparison across 121,370 US tech postings and reported a $33,525 divide, with almost 86% of tech roles paying less for remote. Their stated method is to average onsite pay and remote pay for each job title and compare, broken out by seniority. It contains no control for which employer posted the job, and their own limitations section confirms it.

That is the flaw. Averaging across the market compares Anthropic's onsite jobs against Mozilla's remote ones and reads the difference as a remote penalty.

## Do remote jobs pay less at the same company?

No. Comparing an employer's remote jobs against its own onsite and hybrid jobs at the same level, the average gap is 0.3% in favor of onsite. Half the companies in the sample land between a 7.9% discount and a 5.8% premium, scattered evenly on both sides of zero.

> Across 204 comparisons at 128 companies, the remote pay gap inside a single employer is 0.3%. The market-wide gap is 13%.

The spread across individual employers is wide, so this is not a claim that every company pays identically regardless of work mode. It is a claim that there is no systematic remote discount, and that the market-wide figure cannot be read as one.

## The mechanism has names, and most of them are AI labs

The top of the tech pay distribution right now is frontier AI, and frontier AI does not hire remote.

| Company                                                   | Median advertised | Remote share |
| --------------------------------------------------------- | ----------------- | ------------ |
| [Thinking Machines Lab](https://hotfix.jobs/companies/thinking-machines-lab) | $412,500          | 0.0%         |
| [Anthropic](https://hotfix.jobs/companies/anthropic)                         | $402,500          | 1.1%         |
| [OpenAI](https://hotfix.jobs/companies/openai)                               | $315,000          | 6.5%         |
| [xAI](https://hotfix.jobs/companies/xai)                                     | $310,000          | 8.0%         |

At the other end, [Mozilla](https://hotfix.jobs/companies/mozilla) posts a $149,500 median with every engineering job remote. [Kraken](https://hotfix.jobs/companies/kraken) posts $165,600, also entirely remote. Lumin Digital posts $167,500 the same way.

Drop eight AI labs from the dataset and the mid-level gap falls from 13.3% to 7.7%, while the staff-level gap flips from a 2.2% discount to a 2.0% premium. The senior-level gap does not move, so these companies are a large part of the effect rather than all of it. The broader pattern is the quartile gradient above, and AI labs sit at the sharp end of it.

This connects to something we found earlier: [AI engineering roles carry a 6% to 11% salary premium](https://hotfix.jobs/blog/ai-engineer-salary-2026) over other engineering work at the same level. The employers paying that premium are also the ones least willing to hire outside the office.

## High pay and remote work are not incompatible

The gradient is a sorting effect, not a law, and the exceptions are the most useful names in this analysis for anyone deciding where to apply. [Pinterest](https://hotfix.jobs/companies/pinterest) posts a $270,990 engineering median with 57.0% of jobs remote. [Square](https://hotfix.jobs/companies/square) posts $256,000 at 56.4%. Both clear the median pay of the top quartile while hiring remote at roughly twice that quartile's rate. The same employers show up among [remote senior software engineer roles](https://hotfix.jobs/browse/remote/senior-software-engineer) and [jobs with disclosed salary ranges](https://hotfix.jobs/browse/with-salary).

## Entry-level candidates get squeezed on availability, not price

Remote share of engineering jobs rises with seniority and then falls back at the top:

Chart: Share of engineering jobs posted as remote, by seniority level.

- Entry level: 22.9%
- Mid: 31.3%
- Senior: 37.4%
- Staff: 30.0%

Entry-level engineers face the thinnest remote market and the widest raw gap at once. That compounds a problem we measured in June: [only 1 in 10 tech engineering jobs is open to entry-level candidates](https://hotfix.jobs/blog/entry-level-tech-eng-share-mid-2026) at all. For early-career job seekers, the remote filter cuts an already small market by more than three quarters.

## What this analysis cannot show

These are advertised ranges, not accepted offers. A within-company comparison cannot see a location-adjusted band applied after the offer, so a company posting one national range may still pay a hire in Boise less than a hire in San Francisco. Our finding is about what employers advertise. If location adjustment happens later in the process, it is invisible here.

Worth weighing against that: Stanford's WFH Research puts what workers will trade for two or three remote days at about 5% of pay, with more than half valuing it at 5% or more. If employers wanted to price a remote discount into their jobs, there is a well-measured willingness to pay sitting there. In advertised ranges, they are not taking it.

## Sources

- [JobLeads: Remote vs. On-Site Tech Salaries, The $33K Class Divide (2026)](https://www.jobleads.com/us/blog/job-market-trends/tech-salaries-on-site-vs-remote)
- [WFH Research: Research and Policy (Barrero, Bloom, Davis)](https://wfhresearch.com/research-and-policy/)
- [Barrero, Bloom & Davis: The Evolution of Work from Home (NBER Working Paper 31686)](https://www.nber.org/system/files/working_papers/w31686/w31686.pdf)