Paris vs London for startup founders in 2026

In 2018 this would not have been a real comparison. London was the answer, Paris was a place you visited for an offsite, and the only argument against London was the weather.

In 2026 it is a real comparison. Mistral and the cohort that has formed around it have reshaped the Paris AI scene. The post-Brexit settling has reshaped the London capital scene, mostly upward but in places downward. The cost differential between the two cities has widened to the point where founders are openly choosing.

This is a working comparison. Not aspirational, not theoretical. Where you would actually base, in 2026, depending on what you are building.

What I am comparing

Five things that genuinely matter for a founder choosing a city: capital, talent, the specific scene around your sector, cost, and the events calendar that defines what your week actually looks like.

I am not going to pretend either city is a paradise. They are both serious, expensive places with real frictions. The question is which set of frictions matches your company.

Capital

London still wins on volume. There is more venture capital in London than in any other European city, and at growth stage that gap has not closed. If you are raising a Series B in 2026, you are still going to do most of your meetings in London regardless of where you are based.

Paris wins on early-stage density and on government-backed capital. Bpifrance and the various French state-backed early-stage funds are genuinely unmatched in Europe — the kind of programme that puts a usable cheque on the table for a pre-seed company in a way that no UK equivalent really does. The Family, Hexa, and the rest of the Paris pre-seed ecosystem have matured into something competitive with the London equivalents.

For a seed round, both cities are now real options. For Series B and above, London still pulls.

Talent

London has the deeper senior pool at growth stage. There are simply more people in London who have built and shipped a product to fifty million users, and the network effect of that talent density is real. If you need a VP of Growth who has done it twice, you are hiring in London.

Paris has the deeper research-to-applied talent pool at the AI layer specifically. The combination of ENS, Polytechnique, INRIA, and the post-Mistral alumni cohort creates a density of people who can move from a research paper to a shipped product that London does not currently match. The Anthropic London office and DeepMind both have serious people, but the broader applied-research talent pool is denser in Paris.

At junior product, design, and engineering levels, both cities are reasonable. The London cost of senior talent is, candidly, the single most painful line on a London-based startup's P&L.

The AI scene specifically

Worth a section because it is the question I get most often.

Paris has, in 2026, the most serious AI startup density in Europe outside London. Mistral itself is the obvious anchor, but the company that matters more for the broader scene is the cohort of fifteen to twenty companies founded in 2023–2025 by people who passed through Mistral, the academic labs, or the bigger French AI groups. The events scene around them — the salons at Station F, the dinners in the 11ème, the rotating reading groups at ENS — is small but dense.

London has Anthropic, DeepMind, Google's broader AI presence, and an application-layer scene that is several years more developed than Paris's. The events calendar reflects that — the AI events in London page on Rifio runs at maybe three times the volume of the Paris equivalent in any given week.

Both scenes are real. They are different shapes. Paris is denser at the foundation-model and research-applied edge. London is broader at the application and B2B AI edge. For most AI founders the question is which edge you are working at.

Cost of being there

Paris is cheaper. By a wide margin. Rent in central Paris in 2026 runs roughly 25 to 40 EUR per square metre per month for residential, which is meaningfully under London central equivalents. Senior engineering salaries in Paris run 25 to 40 percent below London for the equivalent role — partly the labour-market pull of London, partly the convenience tax of being there.

The London cost base is not just a line item — it shapes the company. Founders in London raise more, hire faster, burn faster, and hit the next milestone faster. That is the trade. Founders in Paris raise less, hire slower, burn less, and live better. For some companies the first trade is correct. For most early-stage companies, increasingly, the second is.

I will say this carefully because it is the kind of claim people get angry about: the convenience tax in London is real and it has gotten worse. It costs more not just to live but to do anything — the lunch, the taxi, the office, the offsite. By 2026 that compound is meaningful enough to factor into your city choice.

The events calendar

The week-to-week reality of being a founder in either city is shaped by which events you go to. Both cities have real calendars; they look different.

Paris in a typical week: a salon at Station F, a Mistral-adjacent dinner if you are connected, an investor breakfast in the 8ème, a wine night at one of the natural wine bars in the 11ème, and one founder dinner that does the work of three London networking events.

London in a typical week: at least three relevant events any night, the Anthropic London crowd at the AI events in London, a few investor-curated breakfasts, the perennial Soho House founder lunches, and the noise of a much larger calendar to filter through. More signal in absolute terms, lower signal-to-noise ratio.

If you are choosing a city partly on what your evenings will look like, this is a real factor.

Who should go where

A working framework, not a rule:

• Build in Paris if you are AI-first, particularly foundation-model or research-adjacent. If your burn rate matters more than your growth velocity at this stage. If you have any French-speaking advantage, even modest. If quality of life is something you measure.
• Build in London if you are at growth stage and need senior talent at speed. If you are raising a Series B or above and need investor density. If you are selling to global English-speaking customers and need to be where they are. If you are not at a stage where the cost differential dominates your model.

The most interesting founders I know in 2026 are doing both — incorporated in one country, raising from both, hiring across both, and treating the Eurostar as part of their infrastructure. That is increasingly the answer for the founders who can afford to think about it that way.

What changes the calculus

Three things to watch over the next 18 months.

First, whether the AI scene in Paris keeps converting research density into shipped product at the rate it did in 2024–2025. If it does, the Paris case strengthens further. If the cohort slows, London's broader application-layer advantage will reassert.

Second, whether the London cost base flattens or keeps climbing. The convenience tax has been getting worse for five years. If a recession plus a slow softening of the senior-salary market combine, London becomes a more reasonable place to base a 25-person company again.

Third, whether the next generation of European AI capital goes deep at growth stage in Paris or stays concentrated in London. The early signals suggest a slow rebalancing, but slow.

If I had to pick today, with a 30-person AI startup in front of me, I would build in Paris. I would raise some of my next round from London, and I would have an Eurostar pass. That is, increasingly, the answer.

I write about both cities. The Paris this-week page and the equivalent for London on Rifio are the easiest way to keep up with the calendars in both cities — the search filters by city, by category, and by the events that actually matter when you are deciding where to spend a Tuesday evening.

Definately worth keeping both bookmarked if you are running a company in 2026 and undecided.