EU funding for business digitalisation in Lithuania 2026-2027: a guide for SMEs

DESAmedia · 2026-08-24

Short answer

As of late September 2026, there is no open EU funding call in Lithuania dedicated to SME digitalisation. The most recent such call, "Deployment of artificial intelligence solutions" (Dirbtinio intelekto sprendimų diegimas, No. 02-126-K, up to €70,000 and up to 50% of costs), accepted applications until 18 June 2026, and the official list of planned calls does not yet include any new digitalisation calls for businesses. The ILTE "Pokytis" loan for industrial digitalisation remains active, so EU funding for business digitalisation in 2026-2027 will be most useful to companies that prepare their project, indicators and supplier quotes before a call is announced.

Which calls are relevant to digitalisation

In Lithuania's 2021-2027 EU funds investment programme, business digitalisation is funded through progress measure No. 05-001-01-05-05 "Encourage companies to digitalise" (Skatinti įmones skaitmenizuotis) of the Ministry of the Economy and Innovation. The calls are administered by Inovacijų agentūra (Innovation Agency Lithuania). The application is called a project implementation plan (PĮP) and is submitted through the DMS data exchange portal. Calls are published separately for each activity, stay open for several months, and most are competitive: funding goes to the projects with the highest scores.

Below is a summary of the latest calls and other instruments. It shows what size and type of support to expect in future, since new calls usually follow the logic of earlier ones.

On the outlook: on 29 September 2026, the list of planned calls on the EU investment portal (esinvesticijos.lt) contained 57 calls, 6 of which will be administered by Innovation Agency Lithuania. None is aimed at business digitalisation or AI deployment: they are calls for the defence industry, internationalisation and green innovation. Consultants' publications mention tentative dates for new AI calls in late 2026 and early 2027, but these do not yet appear in official sources, so planning around them is risky. The most practical approach is to follow the planned calls list on esinvesticijos.lt and the Innovation Agency Lithuania newsletter.

What counts as eligible costs

Eligible costs are set out in each call's project financing conditions (PFSA), so there is no universal list. The latest calls, however, show a clear direction.

In the AI deployment call (02-126-K), three types of cost were eligible: purchase and deployment of AI products and solutions, rental of software and licences, and indirect costs calculated at a flat rate. The supported activity was defined as deploying AI technologies that allow the company to make better use of its data and digitalise its processes. A general website or a routine upgrade of accounting software does not fall under this definition. A custom solution in which AI plays an essential part, such as document recognition or demand forecasting, may fit the definition, but the final decision rests with the assessors, so borderline cases are worth discussing with Innovation Agency Lithuania consultants before submitting the PĮP.

The AI development call (02-124-K) follows a different logic: it funds the salaries of staff carrying out the project, and external services may not exceed 15% of eligible costs. It is a call for companies that build an AI product themselves rather than buy one from a supplier.

In the digital vouchers call, only services bought from external suppliers at fixed rates were eligible, such as consultancy or preparing a digital innovation deployment plan. The call conditions specifically excluded general consultancy on information that is publicly available free of charge.

Two rules apply almost everywhere: VAT that the company can recover is not eligible, and project activities may not be funded from other public sources, including de minimis aid.

What assessors look for in an application

In a competitive call, assessment means scoring against criteria published in advance. Call 02-126-K had three priority criteria worth 100 points in total, with a minimum threshold of 40 points:

• 40 points if the solution being deployed belongs to one of the AI fields: machine learning, computer vision, smart robotics and automation, natural language processing or digital twins;
• 30 points for the company's high level of digital intensity according to the EU Digital Economy and Society Index (DESI);
• 30 points for the applicant's own contribution.

Beyond the scores, assessors check how well prepared the project is. In the same call, the PĮP had to be accompanied by an annex with detailed, reasoned answers about the project. Without it, the PĮP was rejected without a request for additional documents. The budget had to be supported by commercial quotes from the AI solution developer or an official representative, or by references to market prices, and the own contribution had to be backed by documents, for example a bank's decision to grant a loan.

Monitoring indicators are written into the project contract. If they are not achieved, the provisions of the Project Administration and Financing Rules apply, which may lead to reduced funding. That is why assessors look closely at whether the promised result is realistic and measurable.

Why the project must make sense without funding

In call 02-126-K, support covered up to 50% of eligible costs. The company funds the other half, the ineligible costs and the recoverable VAT itself, and it has to prove its own contribution at the application stage. After the project, licence, maintenance and improvement costs remain, and funding does not cover them.

ILTE (Lithuania's national development institution) says this plainly about its "Pokytis" loans: during assessment, the project must be found economically justified and financially self-sustaining. The same test is worth applying to a grant. If a system only pays for itself when funding covers half its cost, it is probably solving the wrong problem.

There is also a practical reason. A competitive call guarantees nothing: funds may run out, the project may score too few points, and the next call may be announced a year later or on completely different terms. If the project pays off without funding, you can start it with your own money or a loan, and use a call for a later stage. If the project only makes sense with funding, its fate depends on a call whose dates and conditions are unknown.

So our first recommendation is to calculate the return on investment without funding: how many hours, errors or lost orders the current process costs, and how long it will take the system to make up for that. If the numbers work, funding simply shortens the payback period.

Deadlines and what you will not manage if you start late

The application window for call 02-126-K lasted almost 16 weeks: from 27 February to 18 June 2026. That sounds like a lot, but some documents cannot be prepared in that time.

Before a call is announced, it is worth having:

• a description of the process you want to digitalise and its current indicators (time, errors, volumes), because the result will be measured against them;
• commercial quotes from at least a few suppliers with a clear scope, because without them you cannot justify the budget;
• proof of your funding source if you will finance your share with a loan, because a bank's decision may take longer than the call itself;
• approved financial statements for the last two years and a declaration of small or medium-sized enterprise status submitted to the Centre of Registers (Registrų centras);
• a digital maturity self-assessment, if the criteria score digital intensity. EDIH hubs can help with this.

Time matters after the contract is signed too. In call 02-126-K, the project had to start within 3 months of signing the contract, last no more than 12 months and finish no later than 1 September 2029. Building a custom system, including analysis, testing and deployment, only fits into that time if the scope is agreed in advance, so we recommend defining the requirements before applying. We describe how we do this in our process.

The most common mistakes in applications

Starting work too early

In call 02-126-K, the project could only start after the PĮP was registered, and any costs incurred before the funding decision were at the applicant's risk. If the project starts earlier, the whole project becomes ineligible. The start of the project is the first legal commitment to purchase the solution, so a signed contract or a paid advance before registration can cancel the right to funding.

Ignoring procurement rules

Since 1 January 2026, private companies that are not contracting authorities no longer need to publish procurements or prepare procurement conditions. They may also approach a single supplier. However, the price must match the market price, there must be no conflict of interest, a written contract is mandatory when the purchase exceeds €20,000 excluding VAT, and the contract details and a declaration of interests are submitted through DMS. The administering institutions recommend keeping supplier quotes and the technical specification. Buying from a related company, or well above the market price, is a direct route to ineligible costs.

An unjustified budget

An amount not backed by a commercial quote or a comparison of market prices becomes a question during assessment, and sometimes a reason for rejection.

Failing formal requirements

In call 02-126-K, the main activity had to be in industry, and revenue from it had to make up at least 51% of total revenue over two years. Requirements like these are worth checking first, before preparing the PĮP.

Unrealistic indicators

The promised result is written into the contract. It is better to promise less and deliver than to return part of the funding later.

A vague description of the solution

"We will implement artificial intelligence" earns no points. The assessor needs to see which AI field is applied, what data is used and which process will change. If the solution includes integration with existing systems, describe that too. We write about how such projects are put together on our integrations and automation and AI agent development pages.

What to do next

If you are considering a digitalisation project and want to prepare it so that it makes sense both with and without funding, start by describing the process and estimating the scope. We can do this together through IT consulting and system design: the output is a technical specification and estimate that can also serve as the budget justification in an application.

Sources

• Deployment of AI solutions, No. 02-126-K - ES investicijos (in Lithuanian)
• Development of AI solutions, No. 02-124-K - ES investicijos (in Lithuanian)
• Digital SME vouchers, No. 02-042-T - ES investicijos (in Lithuanian)
• SME digitalisation - Innovation Agency Lithuania (in Lithuanian)
• Targeted competence subsidies for business 2026 - Innovation Agency Lithuania (in Lithuanian)
• "Pokytis" loans - ILTE (in Lithuanian)
• Simpler procurement for non-contracting organisations from 2026 - ES investicijos (in Lithuanian)
• Planned calls - ES investicijos (in Lithuanian)
• Digital innovation hubs - Ministry of the Economy and Innovation (in Lithuanian)