French Startup Grants 2026: Bpifrance Aid Shrinks as Rules Tighten

Non-dilutive money is still there, but the margin for error is gone

In the autumn of 2026, French founders hear two different stories about public funding. Founder-focused guides say France is still one of the best places in Europe to combine equity-free grants, R&D tax credits, payroll relief and public loans before giving up much of a cap table.11 The budget figures show a system that has been getting smaller. Bpifrance put €5.2 billion into innovative companies in 2024, and that fell to €3.4 billion in 2025 as the early waves of France 2030 spending wound down.2

Both accounts are true, and the gap between them is the main lesson for founders. The tools still exist. There is less money behind them, and the agencies running them have become more demanding. In practice, a grant now works less like free runway and more like a first, formal review of a company's technology and business case.

What the entry ticket looks like now

The usual starting point is the Bourse French Tech, a Bpifrance grant created in 2014. A detailed accounting guide from September 2026 puts the cap at €30,000. The grant covers no more than 70% of eligible costs and is open to companies registered in France for less than a year with fewer than 50 employees.22 The two limits apply together, so a founder only reaches the full €30,000 with a programme of roughly €42,860 in accepted costs.22 The money comes in two instalments: the first when the agreement is signed, the rest after an end-of-programme report.22

The guides do not agree on the amount. Several founder-oriented roundups describe the grant as €10,000 to €30,000.1214 One innovation-funding overview says it goes up to €50,000.2 Another guide published in June gives a range of €30,000 to €50,000.8 The September breakdown is the most specific. It is tied to Bpifrance's current published conditions and notes that the cap is "subject to conditions", so €30,000 is the safest figure to plan around.22 Founders should take the disagreement as a warning. Program details get repeated and changed as they pass between guides, and one roundup itself tells founders to check every figure with the funding body before applying.23

The guides also disagree about i-Lab, the national competition for founding high-tech companies. Several put the award at up to €600,000.169 Another describes €20,000 plus mentoring.8 The likeliest explanation is that the figures refer to different parts of the programme or different years, but that is a guess. Either way, a founder should not build a financial model on a number taken from a blog.

The ladder above the first grant

For research-heavy teams there are larger options. The Bourse French Tech Emergence is still in Bpifrance's catalogue at up to €90,000 for deeptech projects in companies less than a year old. Spending incurred before the application is filed does not count.22 Further up, the Avance Innovation offers repayable advances of up to €3 million covering up to 70% of costs. The Aide au Développement DeepTech combines grants and advances up to €2 million for advanced R&D and demonstrators. Some guides describe i-Nov as €200,000 to €2.25 million, with co-financing required.12

Tax credits sit underneath all of this. The research tax credit (CIR) covers 30% of eligible research spending up to €100 million a year. Small and medium-sized firms can claim an immediate refund.2 The innovation tax credit (CII) is 20% in mainland France on up to €400,000 of costs and has been extended to the end of 2027.2 The Young Innovative Enterprise status (JEI) still exempts R&D salaries from employer social contributions. However, the corporate-tax exemption was removed for companies created from January 2024, and the status is due to expire at the start of 2029.2

One worked example shows what this stack can add up to. A two-founder quantum-sensor spinout with no revenue combines a French Tech grant with a CIR claim worth about €90,000 in its first year, and reaches roughly €320,000 in public support without dilution before raising any equity.8 The example is a model rather than a typical result. Still, it shows why French deeptech founders often postpone their first priced round.

Why the cuts matter more than the headlines suggest

The 2025 budget changed the rules. It cut France 2030 by about 30%: the plan's 2025 spending fell to €5.2 billion from €7.3 billion in 2024.34 The CII rate dropped from 30% to 20%, and patenting costs no longer count as R&D spending for the research credit.3 The higher CIR deduction for hiring young PhD graduates was also removed.4 France Digitale, the startup lobby, said support in the R&D and pre-commercialisation stages had been reduced.3 In 2025 the group linked the cuts to hiring freezes and some job losses, as fundraising by French startups fell 60% in the first half of that year.1

The reporting differs on how bad this is. A 2025 analysis concluded that French tech could still count on government support. It noted that the prime minister had promised an additional €900 million for startups after earlier signals of cuts.1 The 2026 guides go further, describing roughly €5 billion a year set aside for startup support.11 My reading is that the overall commitment has held but the system now favours different companies. Money is moving towards deeptech, AI and industrial priorities. Bpifrance's 2026–2030 strategy commits €10 billion to AI and breakthrough technologies across all its financing tools.25 Founders building generic software have the least to gain. One roundup said plainly that R&D-heavy startups have better odds than copycat apps.12

Private capital has not filled the gap evenly. French startups have disclosed €9.24 billion across 362 rounds so far in 2026, but much of it is concentrated at the top. Mistral AI's €3 billion Series D and AMI Labs' €890 million seed round lead the table.13 Early-stage teams outside those categories still depend heavily on public money, which is why the shrinking grant budgets affect them most.

Lessons for founders

Size the project honestly. The accounting guide advises scoping a real programme rather than inflating costs to reach the grant ceiling.22 Because of the 70% cap, the company must fund at least 30% itself, through equity, an honour loan or money from friends and family. It also has to cover the costs until the final instalment arrives.22 A founder with no money for that share has a weak application.

Sort out personal financing first. In the Paris region, honour loans from Initiative Île-de-France and Réseau Entreprendre reach €120,000 for technology projects. They are interest-free and require no personal guarantee.17 Securing one before applying to Bpifrance turns a weak application into a fundable one.

Watch the order of steps. Signing quotes before filing can disqualify those costs. The guidance is to meet a Bpifrance account manager before committing to any spending.22 France 2030 applications typically take three to six months to process.2 One guide recommends planning at least a year ahead.8

Track the totals. The EU de minimis ceiling is €300,000 over three years for direct subsidies. Founders who combine several grants can approach it without noticing, while the CIR and CII fall outside it.28 The grant also counts as taxable income in the year it is received. It reduces the base for CIR or CII claims on the same spending.22

Public money as part of a company's track record

For founders who share their progress publicly, there is one more point. The accounting guide argues that the Bourse French Tech matters less for its €30,000 than for what it starts: a first application reviewed by Bpifrance, an ongoing relationship with an account manager, and an innovation-funding history that later funders will look at.22 Investors increasingly treat public validation as evidence, so a grant awarded on a tightly scoped feasibility study is a stronger signal than a polished pitch deck.

The overall picture is that French public funding still rewards founders who prepare, but it no longer covers for those who don't. The grants, credits and loans are still available. With smaller budgets, conflicting published figures and higher selection standards, the founders most likely to benefit treat each application the way a seasoned investor would assess one.