IT startups: which EU funds you can take in 2026, and on what conditions

The mandatory list covers both electronics manufacturing — 2611 manufacture of other electronic components, 2612 manufacture of loaded electronic boards, 2630 manufacture of communication equipment — and software proper: 5821 publishing of computer games, 5829 other software publishing, 6210 computer programming activities, 6220 computer consultancy activities, 6290 other information technology services and 6310 data processing and web page administration. The codes are those of the revised CAEN nomenclature, not the older one.

Beware of a confusion that circulates and can cost an entire file: the same code families appear, in other SME digitalization programmes, on an exclusion list. In POCIDIF Action 2.1 they are the requirement, not the prohibition. Never move a code list from one programme's page to another's — two calls can hold exactly opposite rules for the same numbers.

The figure circulating most on this subject, EUR 3,000,000, is not the ceiling of a software project. It applies to projects whose result is an innovative hardware product manufactured by the beneficiary, and the part above EUR 1,500,000 must be justified by financing the production line. For an application, a service or a platform the ceiling is EUR 1,500,000. A startup budgeting from three million mis-sizes its project on day one.

The guide also sets tiers for aid granted to newly established enterprises: EUR 750,000 in the more developed region, EUR 1,000,000 in the less developed regions and up to EUR 1,500,000 for innovative newly established enterprises, in all cases within the maximum per-project ceilings. The call allocation is EUR 48,813,278 from ERDF and the state budget.

The guide explicitly forbids outsourcing all technical activities. The beneficiary must demonstrate direct involvement and its own coordination capacity, and the non-repayable aid allocated to technical activity carried out by its own staff must represent at least 20% of the aid allocated to the core activity. The implementation team must also include at least one technical coordinator and two technical experts qualified and experienced in the project's field.

The consequence for a small startup team is direct: we cannot be contracted as your entire technical team. We can cover components, stages or peaks, but the technical core must stay with you, and those people must exist at submission rather than in intention. We say so before signing, because it is the only version in which the contract survives verification.

The guide also carries a rule that bears directly on the supplier–consultant relationship: costs where there is suspicion of accommodation offers between the applicant and the companies that issued offers, or those that produced the documentation for submitting the project, are treated as ineligible and the project scores zero on that sub-criterion. The offering company's field of activity is checked against what it offered. Separating whoever writes the dossier from whoever delivers the software is not our preference; it is a requirement of the call.

The call's indicators are few and sharp: one innovative product, application or service delivered, and three proofs of commercialisation. The proofs can be concluded contracts or invoices; proforma invoices and agreements do not count as proof of commercialisation. To these is added investment performance, which must exceed 0.5, calculated as direct and indirect revenue from the developed product divided by its development costs, measured at the end of the three-year durability period.

On scoring, approval requires a minimum of 70 points; below that, projects are rejected. The excellence threshold is 92 points, and projects reaching it enter contracting in submission order up to 50% of the call allocation. For a very strong project, in other words, the submission date decides the place in the queue. The grid also penalises two things you control from the budget: a hardware share above 20% of total eligible value on software-result projects, and costs declared ineligible at evaluation above 15% of those requested — both score zero on the budget criterion.

Start-Up Nation session 2 plays in another league: smaller sums, but accessible to freshly incorporated companies, with a business plan instead of an innovative-product dossier. The current period, with no submission dates published, is useful precisely because it allows the digital chapter to be built on real specifications and costs rather than hearsay estimates.

An eligibility warning that is rarely written down: the EDIH network is not a call a company applies to. The applicant is the hub, and the company receives support through the hub in the form of non-financial state aid granted under a contract — services, testing, digital maturity assessment, not money in the account. Any page inviting you to submit a project to an EDIH describes the mechanism incorrectly.

We are not funding consultants and we do not submit files. We deliver the technical component the evaluation reads: specifications, architecture, development schedule, a budget by eligible cost category and a draft contract conditional on obtaining the funding. In execution we can act as an external team on components, within the limits described above.

We promise no scores, we issue no eligibility verdicts and we do not write the application. If the product you want to budget cannot reach the call's indicators within the implementation period, we say so before submission — including when that means we do not sign the contract.