FDE pay in Paris: why €205,000 and €60,000 cannot be compared

Two Forward Deployed Engineer postings in Paris. Anthropic’s lists a base salary of €205,000–220,000 a year. Leadbay’s, from a Paris startup, lists €50,000–70,000. Side by side, one is roughly three to four times the other, and it is easy to conclude that to be an FDE in Paris you need to get into a lab.

That conclusion is hasty, and it is worth seeing exactly where it goes wrong. The two figures were recorded at different times, the packages are structured differently, and the experience required differs too.

If you are weighing an FDE offer in Europe, or learning to read job postings to choose a way into the field, the method for pulling these two numbers apart is worth more than the numbers themselves.

A salary range only tells you something when you know when it was recorded, what it includes, and what level of person it pays.

The €205,000 figure is older than you think

Anthropic’s “Forward Deployed Engineer, Applied AI” posting in Paris survives as a copy on Accel’s job board, and that copy clearly states the role is no longer accepting applications. It was posted more than six months ago. The €205,000–220,000 range is therefore data from early 2026, not an offer you can still apply for.

An analysis of Anthropic’s FDE salary ranges, published on 20 October 2026, shows the current picture is even harder to read. According to it, the US range is $280,000–320,000 and the London range is £225,000–255,000.

Among the postings the analysis reviewed at the time, neither Paris nor Munich disclosed a range. The Paris data is thus both thin and stale: a single data point, taken from a closed posting.

That point sits below the US range, but resist converting currencies and computing a percentage. Perspective’s 2026 FDE salary report, which surveyed 1,200 FDEs, offers a rule of thumb: ranges outside the US typically come in at 50–70% of US ranges.

The report itself says its data is US-only, so the rule is for reference. It tells you roughly how wide the gap is, not what any particular person should be paid.

Base salary is not the whole package

The harder part is equity. Anthropic’s posting describes the job as working inside the customer’s organisation to deliver AI applications.

On compensation, it mentions only competitive pay and an optional programme: the company adds a contribution when employees donate equity (equity donation matching). It does not state any specific equity grant.

Perspective’s report, meanwhile, says that in 2026 equity makes up 60–70% of total compensation at frontier labs, up from 35–45% two years earlier.

If you compare only base salary between a lab and a startup, you are comparing a small part of one package with nearly all of the other.

So when negotiating with a lab, the first question should be how the equity is calculated, not what the base salary is.

Leadbay is the opposite: its posting states equity up front, 0.1–0.5%. But the same report says applied AI startups pay markedly lower total compensation than frontier labs. Its advice is blunt: plan on 5–7 years before you can sell the equity.

Until then, 0.5% equity will not pay rent in Paris.

French gross pay is not what lands in your account

The third layer is the one engineers from other markets most often miss. According to the 2026 URSSAF social contribution rates compiled by Legisocial, the employee’s share is about 20–23% of gross pay. The employer’s additional share can reach about 40%.

Apply that to the middle of Leadbay’s range. At €60,000 gross, subtracting 20–23% leaves about €46,200–48,000 before income tax. On the company’s side, if contributions reach 40%, the cost of the role could be around €84,000.

This is an illustration using average rates, not a real payslip, since actual rates vary case by case. But it shows that the same gross figure means three different things depending on where you stand.

When comparing a Paris offer with one in London or the US, then, do not set gross against gross.

These are really two different jobs

The last layer has nothing to do with money. Leadbay’s posting accepts people with no experience and requires customer visits of at least about 6 days a month, sometimes up to 3 days a week. The role is for people new to the field, who will learn by being on site with customers very often.

For Anthropic’s posting, the available data does not show the experience requirement. What is certain is only that it is a role delivering AI applications inside the customer’s organisation, at a company whose FDE ranges in the US and London are both above two hundred thousand.

When one posting explicitly takes newcomers and the other pays three to four times as much for the same title, the two are most likely hiring different kinds of people. Calling the entire gap a difference between two markets is a hasty conclusion.

Paris’s overall pay level matters too. A comparison in L’ADN found Paris had the lowest average tech salary of the 16 regions surveyed, about €49,000, against about $134,000 in Silicon Valley.

The article attributes the gap to market maturity and cost of living. Against that baseline, Leadbay’s €50,000–70,000 is by no means low for a role open to people without experience.

If you are new to the field, which should you choose?

If you have just graduated, or have written code for a few years but never worked directly with customers, treat a role like Leadbay’s as an investment, not simply as a low salary.

You are paid around the average for tech in Paris without ever having worked as an FDE. Every month you spend at least about 6 days alongside customers. That is on-site experience you can later describe with concrete numbers.

For someone with a few years as an engineer, the trade-off needs more care: you may accept a newcomer’s salary in exchange for the experience you lack. It is worth doing if that gap is exactly what keeps you out of better-paid FDE roles; otherwise, look for a role that matches your current level.

Either way, when writing your CV, do not just write “worked with customers”. State how many times you went on site, what you delivered and how the customer used it.

When an FDE role is described as delivering AI applications inside the customer’s organisation, concrete evidence that you have done that work persuades more than a list of technologies.

When reading any FDE posting, in Paris or elsewhere, check five things before looking at the salary. Is the posting still open, and when was it published? Is the figure gross or net? How much equity is there, and when can it be sold? How many days of customer visits are required? Does it accept newcomers?

€205,000 in a closed posting and €60,000 with 0.5% equity that cannot yet be sold cannot be measured with the same ruler.

Public data on the Paris FDE market is still too scarce for anyone to give a “standard” range. While waiting for more postings to disclose pay, the person who can split a figure into its layers will negotiate better than the one who has merely memorised it.