If you’re a founder reading this because you’re weighing whether a French incubator or accelerator makes sense for you, I want to save you the frustration I’ve watched other founders live through.
Most of the “best accelerators in France” articles you’ll find online were written years ago and have quietly gone stale. They still list programs that have shut down. They still list programs that stopped investing. In one case, they still list a program whose co-founder is currently being pursued by French, British, and Cayman Islands courts on charges of embezzling €4.5 million from the very fund those founders were joining. I wish I were making that up.
So before I get to who’s genuinely worth your time in 2026, I want to be upfront about how I’ve written this list.
Every program below is currently operating, and I’ve verified that as of publication. Every one gets the same treatment: what it actually is, what founders it’s built for, real success numbers where the programme publishes them, a plain “figures not disclosed” note where it doesn’t, a real named alumnus, and one fact-checked limitation. Every entry gets a limitation, without exception, because a list without them isn’t a guide. It’s a brochure.
This is not a sponsored article. Neither Altar.io nor I receive a cent from any of the programs listed. And a note on our vantage point: at Altar we help founders build the products they take into accelerators, and we work with founders who come out of them. That means we’ve seen what happens on both sides of these programs’ doors, and we can offer that perspective. Then, if you need some help with the actual product, we can talk about that too.
One last thing before the list. The single most useful piece of advice I can give you about picking a French program isn’t in the list itself. It’s this: the French ecosystem has turned over more in the last three years than it did in the ten before that. Techstars paused its Paris accelerator in June 2025 and hasn’t resumed. Entrepreneur First closed its Paris office in 2025 to focus on the US. The Family, once one of France’s most famous names in this category, is no longer investing and is deep in litigation. Whoever wrote the listicle you read before this one probably didn’t check. Please check.
First: incubator, accelerator, or campus? They are not the same thing
I get this question in almost every call I take with a first-time founder, so let me answer it once, cleanly, before we get to the list.
An incubator takes you very early. Sometimes with just an idea and a co-founder, and it helps you build. Long timelines, 12–36 months is common. Often modest or no equity taken. Usually attached to a university, a research institution, or a public body. Agoranov is the archetype. You don’t leave with a huge investor Rolodex. You leave with a defensible company.
An accelerator takes you when you have something working, or nearly working. Short timelines, usually 3–6 months. A defined batch, an equity stake or investment. The trade is speed and network for a piece of your cap table. Y Combinator is the archetype; Wilco and Founders Factory Paris are the closest French equivalents.
A startup campus is the newer category, and mostly it means Station F. It isn’t a program itself. It’s a physical and organisational structure that houses dozens of programs from different partners under one roof. Getting into “Station F” means getting into one of those programs. Not into some monolithic thing called Station F. I’ll explain why that distinction matters when we get there.
Three different tools. Different founders need different tools. If you tell me “I have an idea and two co-founders and no money,” we should be talking about an incubator. If you tell me “I have paying users and I need to raise a seed,” we should be talking about an accelerator. Pretending they’re the same thing is one of the most common and most expensive mistakes I see!
The list, on the record
I’ve grouped these by what they’re actually best at, not by ranking. A rank order would be a lie. The “best” one for a healthtech founder is a bad fit for a fintech founder. And no fair guide can pretend otherwise.
Station F, the campus, not the program
Let me start with the one that isn’t really a category.
Station F is not an accelerator. It’s a campus in Paris’s 13ème arrondissement, in a repurposed rail freight hall. It houses more than 40 different programmes from partners as varied as Meta, Microsoft, LVMH, Sanofi, BNP Paribas, and Doctolib. It’s the largest startup campus in the world by any reasonable definition. It has hosted thousands of startups since opening in 2017. It’s directed by Roxanne Varza, who took the role at 31 and has run it since. When founders say “I got into Station F,” what they actually mean is “I got into one of the 40+ programmes based at Station F.” F/ai for AI. Fighters for unconventional founders without a degree. HEC’s incubator, INSEAD’s Launchpad, Meta and Hugging Face’s open-source AI programme, Microsoft’s GenAI accelerator with Mistral and Nvidia, LVMH’s La Maison des Startups for luxury, Sanofi’s for pharma AI, and dozens more!
That matters because “how do I get into Station F” is the wrong question. The right question is “which of the 40+ programmes at Station F fits my company.” Get that right and Station F is the highest-density place to build a startup in Europe. You’ll walk past founders of unicorns getting coffee. Get it wrong, and you’ll waste six months in a programme built for someone who isn’t you.
Fact-checked limitation: the density is also the noise. Everyone in Paris tech wants to talk to Station F founders, which means calendar chaos and a permanent temptation to network instead of build. Varza has said publicly that the ecosystem’s biggest problem is the exit market. “The exit market is the big issue,” she told Sifted. If you’re building a company hoping for a European exit rather than transatlantic, that’s a real constraint the campus can’t solve for you. The other one: Paris is expensive, and Station F does not solve your rent problem.
Best for: founders who already know which of the 40+ programs they belong in, and want the density of the world’s biggest startup campus around them while they build.
Agoranov, the deeptech incubator behind Doctolib, Alan, and Criteo
If Station F is the campus, Agoranov is what a serious incubator looks like when it doesn’t need to be a campus to be world-class.
Agoranov is a not-for-profit incubator founded in 2000 by four Paris universities and a research institute. It has quietly produced one of the strongest alumni rolls in European tech. Doctolib, Alan, Criteo, Dataiku, Ynsect, Shift Technology, and hundreds more. Doctolib’s founders, Stanislas Niox-Chateau and his three co-founders, were incubated by Agoranov in 2013, in the earliest days of what is now a multi-billion-euro company. Alan came through. So did Criteo, one of the very few French tech IPOs on Nasdaq to have crossed real scale!
The Agoranov model is 12–24 months of support, dedicated office space in Paris, coaching, and access to public funding. Not equity. That last point matters. Agoranov does not take a stake. It’s public-mission by design, which changes the incentives entirely. Roughly 450+ startups have been through the programme cumulatively. Agoranov is unusually strong in deeptech and life sciences specifically because of who founded it.
Fact-checked limitation: it’s slow. This is not the programme for a founder in a hurry, or a founder building a consumer product with a six-month window to prove traction. It’s for a founder who has a real technology and needs the space and support to make it a real company. If you’re not scientific or engineering-heavy, you’re probably better served elsewhere.
Best for: deeptech, healthtech, and hard-science founders who need incubation over acceleration.
Wilco, no equity, no batch, and 50+ exits nobody talks about enough
Wilco is the most under-marketed serious accelerator in France, and it’s the one I recommend most often to founders who ask me privately.
Founded in 1999 as Scientipôle Initiative and rebranded to Wilco in 2015, it has accelerated more than 1,200 startups. Alumni include BlaBlaCar, Dataiku, Ynsect, ManoMano, and Sigfox. The model is genuinely unusual. Wilco does not take equity. It offers interest-free loans of up to €150,000, repaid over 5–7 years, and structures its programme around a hard target of €1M ARR within three years. It’s the only French programme I know of that ties its success explicitly to the founder’s revenue milestone rather than to a demo day. More than 50 of its alumni have had exits!
The “no equity” point deserves emphasis. For a French or European founder, the standard accelerator trade is 5–10% of your company for €100–150k and a network. That’s a bad deal if you’re already coachable and already have some traction. Wilco gives you the same money without the cap table hit. There’s a reason BlaBlaCar came through Wilco rather than Y Combinator. If you want the same low-friction terms from a build partner, our own paid scoping phase is designed the same way. You commit to one small paid engagement, and you can walk away with no lock-in.
Fact-checked limitation: the interest-free loan is repayable, which is a very different psychological arrangement than an equity investment you can walk away from. If your company fails, you and your co-founders are on the hook to return the money over time. That is a real constraint on risk-taking. Also, Wilco is much less known outside France. The international network you’d build in a YC batch is not what you get here. The network is deep and French.
Best for: early-revenue founders who are French or French-adjacent, want capital without dilution, and are prepared to commit to a real revenue milestone.
50 Partners, the peer-to-peer model, backed by the founders of BlaBlaCar and Aircall
50 Partners is small on purpose. Each year, only 6–8 startups are selected out of more than 2,000 applications. The selection ratio makes it one of the most competitive accelerators in Europe by that measure alone. The “50 partners” the name refers to are 50 successful French entrepreneurs. The founders of BlaBlaCar, Aircall, Dataiku, Zenly, PayFit, and others, who each commit to mentoring the batch.
The model is “lifelong support” rather than a fixed programme. That sounds like marketing language until you talk to alumni. Portfolio companies include DataDome in cybersecurity, SimpliField in retail SaaS, and dozens of others that have raised follow-on rounds. 50 Partners takes 5% equity for its support and a small investment.
The thing that makes 50 Partners genuinely different is the density of operator wisdom in the mentor pool. When a BlaBlaCar co-founder is one of the 50 who has committed to responding to your text messages, you’re getting something you cannot buy any other way.
Fact-checked limitation: 6–8 startups a year means most founders reading this will not get in. The application process is intense. And the “50 partners” are almost entirely French, which is a strength for building in France and a limitation if your target market is the US or Asia and you need mentors who’ve built there.
Best for: European founders, especially French, who want the deepest possible operator mentor network and don’t mind giving up 5% for it.
Founders Factory Paris, the corporate-backed accelerator with L’Oréal and Aviva
Founders Factory is a London-headquartered venture builder and accelerator that opened its Paris office in 2019. Its Paris presence is unusual for one reason. It runs corporate accelerator programmes in partnership with L’Oréal and Aviva. The L’Oréal Beauty Tech accelerator has selected multiple cohorts of beauty-tech startups since 2019. The Aviva fintech programme has done the same for financial services and insurance.
The model is a hybrid. Founders Factory acts as both an accelerator, with six-month programmes and equity investment, and an incubator, venture-building startups from scratch inside the studio. Companies that go through the L’Oréal programme, for example, get direct access to L’Oréal’s beauty industry network. For a beauty startup, that’s worth more than most funding rounds.
Fact-checked limitation: the corporate-backed model means your programme is genuinely built around L’Oréal’s or Aviva’s strategic interests, not yours. If you’re a beauty tech founder whose vision is aligned with L’Oréal’s, that’s an accelerant. If it isn’t, there will be friction you don’t want. Also, Founders Factory takes a meaningful equity stake. This is not a low-dilution option.
Best for: beauty tech founders (L’Oréal) or fintech/insurtech founders (Aviva) whose vision aligns with a major corporate partner’s strategic interests.
Le Village by CA, the corporate accelerator network with a real footprint
Le Village by CA, launched by Crédit Agricole in 2014, is a network of 44+ Villages across France, Italy, and Luxembourg that collectively host around 800 startups. The Paris Village is the flagship. It’s a 23–24 month acceleration programme with a heavy focus on connecting startups to Crédit Agricole’s corporate partners.
It’s not equity-driven. Le Village charges startups rent for their space and offers services around that. What you’re really buying is proximity to Crédit Agricole’s enterprise clients. For a B2B fintech or a company building anything the bank might buy, that’s a real value.
Fact-checked limitation: Le Village is best for startups selling to enterprises Crédit Agricole banks. If your business model doesn’t have a natural bridge to the bank’s client base, you’re paying for a network you’re not using. And unlike Wilco or 50 Partners, Le Village does not curate for scale-up potential. You’ll be next to companies at very different stages.
Best for: B2B startups (especially fintech, agritech, insurtech) whose target customers are the kinds of companies Crédit Agricole banks.
Bpifrance’s programmes, the state-backed option that too many founders skip
Bpifrance is the French public investment bank, and it runs, directly or as a co-financier, a large share of the funding available to French startups. Bpifrance’s role is different from an accelerator’s. It doesn’t run a fixed cohort programme in the way Wilco or 50 Partners do. What it does is provide grants, loans, and equity investment through vehicles like French Tech Emergence, French Tech Tremplin, and the Deeptech Plan.
I’m including Bpifrance in this list because too many first-time French founders don’t realise how much of the French tech ecosystem’s growth was built on Bpifrance capital. And because for a large class of founders, deep tech, hardware, anything requiring capex before revenue, Bpifrance’s non-dilutive funding is the difference between existing and not existing.
Fact-checked limitation: Bpifrance funding is French. If your company is not French-domiciled or does not have a strong French link, most of the vehicles won’t be open to you. And the paperwork is famously heavy. Bpifrance is a public bank, not a startup studio, and it moves at public-bank speed. Expect the process to take months, not weeks.
Best for: French-domiciled deeptech, hardware, and capex-heavy founders looking for non-dilutive capital.
A word on the wave of AI-specific programmes at Station F
If you’re an AI founder in 2026, France has become one of the two or three most credible places in the world to build. That’s not marketing. It’s a function of Mistral being French, the government’s aggressive AI positioning, and the specific programmes now clustered at Station F.
The Meta and Hugging Face open-source AI programme, the Microsoft GenAI accelerator run with Mistral and Nvidia, and Station F’s own F/ai programme are three of the strongest AI-specific accelerator programmes anywhere in Europe. For a founder building foundation-model-adjacent AI, or applied AI where being close to Hugging Face’s ecosystem matters, this cluster is genuinely differentiated!
That said, I want to flag the same warning I flagged in a separate article about AI product agencies in Europe. Beware anyone who answers “we’ll just use GPT for that” to every question. The best AI-specific programmes at Station F are the ones that push their startups toward defensible technology, not toward wrapping a language model in a UI. Ask questions about that in your interview.
How to actually choose, in four blunt questions
I gave a version of this list of questions in my agencies article, and it works just as well here.
Does the programme’s model match your stage? If you have an idea, you need an incubator. If you have traction, you need an accelerator. If you have neither, you probably need to keep building before you apply anywhere.
Does the programme take equity, and is the equity a fair trade for what it gives you? If it takes 7% for €100k and a mentor Rolodex, you’re paying €14k per point of your company. Ask whether the network is genuinely worth that. Wilco and Agoranov not taking equity is not a coincidence. It’s a design choice with real implications.
Can they show you real alumni you can call? Every programme in this list will introduce you to alumni if you ask. If they won’t, that’s your answer. The alumni who won’t take your call are more informative than the ones who will.
Will the programme still exist in three years? This is the question nobody asks, and it’s the one Techstars Paris founders wish they’d asked in 2024. Look at how long the programme has been running, who’s backing it, and whether it’s raised its own recent fund. Programmes that depend on a single corporate partner or a single institutional fund are more fragile than they look.
Do you have a brilliant startup idea that you want to bring to life?
From the product and business reasoning to streamlining your MVP to the most important features, our team of product experts and ex-startup founders can help you bring your vision to life.
The eight, at a glance
If you want the whole picture on one screen, here it is.
| Programme | Type | Location | Equity taken | Programme length | Best fit for |
|---|---|---|---|---|---|
| Station F | Campus (40+ programmes) | Paris 13ème | Varies by programme | Varies | Founders who know which of the 40+ programmes they fit |
| Agoranov | Incubator | Paris | None | 12–24 months | Deeptech, healthtech, hard-science founders |
| Wilco | Accelerator (loan-based) | Paris + national | None (repayable loan) | ~3 years to €1M ARR | Early-revenue French founders wanting capital without dilution |
| 50 Partners | Accelerator | Paris | 5% + small investment | “Lifelong” | Founders who want the deepest French operator mentor network |
| Founders Factory Paris | Accelerator + venture builder | Paris | Meaningful stake | ~6 months | Beauty tech (L’Oréal) or fintech (Aviva) founders aligned with corporate partners |
| Le Village by CA | Corporate accelerator | Paris + 44 locations | None (rent-based) | 23–24 months | B2B fintech, agritech, insurtech selling to Crédit Agricole clients |
| Bpifrance programmes | Public funding vehicles | National | Varies | Months to years | French deeptech, hardware, capex-heavy founders wanting non-dilutive capital |
| Meta/Microsoft AI programmes at Station F | AI-specific accelerators | Paris 13ème (Station F) | Varies by programme | 6–12 months | Foundation-model-adjacent AI founders |
Numbers verified live at the time of writing. Take a screenshot; specifics will shift, but the picture holds.
And what’s missing on purpose
You’ll notice I haven’t included Techstars Paris, Entrepreneur First Paris, or The Family. That’s not an accident.
Techstars Paris is paused since June 2025 and, per Techstars’ own statements, is not expected to return before 2026 at the earliest. Applying now means applying to a programme that doesn’t exist. If it returns, I’ll add it in the next update.
Entrepreneur First’s Paris office closed in 2025 as EF pivoted to focus on the US. The Hexa entity that replaced it is a fund, not an accelerator. If someone recommends EF Paris to you in 2026, they’re reading an old article.
The Family, which appears on every listicle written before 2022, hasn’t actively invested since 2020. Its co-founders are managing what remains of its portfolio while pursuing its third co-founder across three jurisdictions on charges of embezzling €4.5 million. In February 2025, that co-founder was arrested at Nice airport. Whatever the outcome of the legal proceedings, and while nothing has been proven in court, The Family in 2026 is not a functioning accelerator, and I’d be doing you a disservice to list it as though it were.
Every one of those names still appears on the top ten “best French accelerators” articles on the internet as of this writing. Please cross-check.
So, where does that leave you?
If you’re building deep tech in France, the answer is probably Agoranov, possibly followed by a Bpifrance programme for non-dilutive scale-up capital. If you’re an early-revenue B2B founder who’s French or French-adjacent, Wilco is genuinely under-rated and I’d apply there before anywhere else. If you have an AI-heavy vision and want to sit next to Mistral and Hugging Face, the Meta or Microsoft programmes at Station F are the strongest cluster in Europe. If you’re a beauty-tech or fintech founder whose vision aligns with L’Oréal’s or Aviva’s, Founders Factory Paris. If you want the deepest French operator mentor network and are ready for a very competitive selection, 50 Partners.
The one thing every one of these programmes has in common. They will change what you build, and you should want that! If you apply because you want the logo on your deck, you’ll be disappointed. If you apply because you want feedback that will genuinely change your product, you’ll get more than your money’s worth.
Good luck out there!
If you’d like a second pair of eyes on the product they’ll be evaluating when you get in front of one, that’s the kind of conversation we have most weeks at Altar.io.
Schedule a Call and you’ll be talking to product and tech experts, with real experience doing these things.
Thanks for reading.
