TL;DR. Since July 7, 2026, ZIM, Germany’s Central Innovation Programme for SMEs, has not accepted new applications or project outlines. Resumption is targeted for early 2027, depending on the 2027 federal budget. By November 3 we had found no announcement of a reopening. Use the pause to prepare: whoever has a finished outline ready when it reopens submits in the first wave. On substance, nothing has changed: “technological novelty” does not mean a world-first invention, but a demonstrable leap in development for your company with real technical risk. Success depends above all on the technical approach.
Update: application freeze since July 2026, and how to use the time
Since July 7, 2026, 12 noon, a temporary freeze on applications has applied to all ZIM project types and project outlines. That is what the Federal Ministry for Economic Affairs announced, as published for example by the Regensburg Chamber of Commerce. The reason is the sharp rise in demand since the fourth quarter of 2025 against limited budget funds.
- Exception: applications with international partners under open bilateral or multilateral calls are still accepted. If you have a suitable partner abroad, that route remains open.
- Not affected: applications already submitted will be approved as long as they meet the conditions and funds are available. Payments for approved projects continue as normal.
- Resumption: targeted for early 2027. The decision depends on the funds in the 2027 federal budget; there is no firm commitment. Future funding conditions and updated forms are to be published on zim.de. So expect that details may change.
Our advice: treat the freeze as preparation time, not waiting time. When demand builds up, those who are ready win at the reopening: with a finished outline, access to the application portal set up and complete documents on creditworthiness. Concretely: research the state of the art, sharpen the technical approach, calculate the work and exploitation plans, gather the accounting documents. Everything else in this post still applies unchanged; only the time of submission shifts.
What a ZIM application is, and what it is not
ZIM applications are systematically underestimated or overestimated. Underestimated by those who think you send a PDF with an idea and receive money. Overestimated by those who do not even try without a research professor. Neither is true.
The Central Innovation Programme for SMEs is Germany’s large federal funding programme for research and development in small and medium-sized companies, open to all topics. A new guideline has applied since January 1, 2025. Eligible are companies with fewer than 500 employees and at most €50 million in turnover or €43 million in balance sheet total. In cooperation with an SME, companies with fewer than 1,000 employees may also take part. Eligible costs are capped:
Eligible costs under the 2025 ZIM guideline, caps
Grant for a fully used single project: €172,500 to €310,500 at a funding rate of 25 to 45%
The funding rate for single projects is between 25 and 45%, depending on company size and region. For cooperation projects it is higher, according to advisory sites up to 60%. A substantial grant, realistically achievable if the application is right. This post shows how we write ZIM applications: with the phrasing that convinces reviewers, and the phrasing that blocks them.
What ZIM really funds
The often-quoted “technological novelty” is the most common misunderstanding. It does not mean you have to invent something the world does not know. It means the project represents a technological leap for your company and that the risk can be documented. Fundable projects meet three criteria:
- Technological risk. Not “we implement a known solution”, but “whether it is feasible only becomes clear during the project”.
- Innovation content. The project advances the state of the art, not necessarily worldwide, but for the industry or the use case.
- Commercial exploitation. The end result is marketable. An internally used method counts too if it measurably improves competitiveness.
AI projects in ZIM: what counts as technological novelty
ZIM is open to all technologies and industries, and many outlines now revolve around AI. But AI projects are judged by the same criteria as any other, and that is exactly where many fail.
Fundable patterns:
- New or substantially further developed algorithms. Not the use of a finished model, but its methodical development for a specific problem.
- Applied AI systems with technical risk. For example, when it is open whether the available data are sufficient for the use case or whether the required accuracy is achievable in real operation.
The knockout criterion is introducing standard software. Anyone who introduces, configures or connects an available AI solution via an interface does not advance the state of the art. There is no ZIM for that, however useful the project is for the company.
Two typical patterns show where the line runs. A machine builder wants to detect faults on production machines early, but there are no ready-made models for its machine data: the novelty lies in transferring machine learning to these data. Or a company wants to optimise machine controls with AI: the technical risk lies in whether the optimisation holds under real production conditions. In both cases: not “we use AI”, but a named, open technical problem.
The outline: what has to go in
The project outline is optional; you can go straight to the full application. We still recommend it almost always. It gives you a solid assessment of fundability from the responsible project agency before the effort of the full application. Five sections also form the framework of the full application:
- Starting point and motivation. Why this project now? The concrete trigger: competitive pressure, changed market requirements, a technological gap. Short and without marketing language. Anyone who opens with “a strategic imperative” has lost by the third sentence.
- State of the art. What solutions exist, and why are they not enough? This shows that you know the market and which gap your project closes. Without this comparison, the innovation content cannot be proven.
- Technical approach and innovation content. Which method, which architecture, which algorithms? The more concrete, the better. “We use AI” is not enough. “We are developing a multi-stage search combining keyword and vector methods for the technical language of mechanical engineering” convinces.
- Work plan and schedule. Milestones, work packages, resources, realistic rather than optimistic. A plan that crams all development into the first months looks implausible. Credible plans have a calm start (analysis, setup, prototype) and an intense middle.
- Exploitation plan. Internally: how do costs or competitive position change? Externally: partnerships, licences, products? The more concrete, the more credible.
The phrases that stop reviewers
We delete these sentences from outlines:
- “The AI transformation of SMEs is an urgent topic.” Says nothing about the project.
- “We rely on state-of-the-art AI technologies.” Which ones exactly?
- “The project has enormous economic potential.” Quantify it or leave it out.
- “The results will be used.” Passive and non-committal.
What works instead, as example phrasing:
- “Unlike existing approaches (reference A, B), our approach solves problem X with method Y, used in this context for the first time.”
- “The technological risk lies in the missing data basis for the technical terminology. The project will test whether synthetically generated data close this gap.”
- “If development succeeds, processing time for a quote calculation drops from 4 hours to an estimated 40 minutes. At 80 calculations a month, that equals about 1.7 full-time positions.”
The full application
If the assessment of the outline is positive, or if you skip the outline, the full application follows. It is submitted digitally and deepens every point of the outline. Added to that:
- A detailed financial plan, whose costs follow traceably from the work plan.
- Documents on creditworthiness: balance sheets, management accounts, annual financial statements of recent years. The review checks whether you can carry your own share.
- Declarations on de minimis aid and other state aid rules.
The most common delay is incomplete financial documents. Gathering them at the outline stage saves weeks later. This is exactly what can be prepared well during the freeze.
Frequently asked questions about the ZIM application
Is ZIM currently open for applications?
No. Since July 7, 2026, a temporary freeze has applied to all project types and to outlines too. Applications with international partners under open calls are exempt. Resumption is targeted for early 2027.
How high is the ZIM funding rate?
For single projects 25 to 45%, depending on company size and region; more for cooperation projects. Eligible costs are capped at €690,000 per single project and €560,000 per company in a cooperation. Whether the conditions will stay the same after the reopening is open.
How long does approval take?
That depends on the volume of applications and on follow-up requests. The biggest lever in your own hands is the completeness of your documents.
Can I combine ZIM with Germany’s research tax allowance?
Yes, but the same cost item may never be funded twice. The combination becomes practical with cleanly separated work packages and cost items, documented from the start of the project.
Who may apply for ZIM?
Companies with fewer than 500 employees and at most €50 million in turnover or €43 million in balance sheet total; in cooperation with an SME, also companies with fewer than 1,000 employees. Research institutions can take part as cooperation partners.
Do I need a project outline?
No, the outline is optional, but a sensible preliminary check. During the freeze, the project agency does not accept outlines either.
How long it takes, realistically
A ZIM single-project application goes through these steps:
- Prepare and submit the outline.
- Assessment of the outline by the project agency.
- If positive: prepare and submit the full application.
- Review and decision.
From idea to decision, experience shows it takes several months. Plan for that and you lose no time. Wait for the decision before starting the project and you lose almost a year. And if you use the months until reopening for step 1, you are in the first wave.
As of November 3, 2026, freeze taken into account. ZIM conditions are adjusted regularly. If you want to check an AI project for fundability, we are happy to talk.
