A deep tech company applying for funding has a problem most startups don't have: no usage curve, no revenue, sometimes no working prototype. The product might be eighteen months from existing at all. Every programme that genuinely funds this stage has to substitute something for traction, and that something is technical evidence: has the science cleared a specific, checkable milestone, does the team hold the credentials to clear the next one, and does the money address a real non-scientific risk (manufacturing, regulation, a second round) rather than just general operating cost. We track 48 programmes open to science and deep tech companies out of 186 in our catalogue, roughly one in four, and the eligibility text usually tells you in the first two paragraphs whether a programme was built for this or just doesn't exclude it.
Built for this, or merely not against it
The tell is specificity about stage and sector, not a generous dollar figure. A programme that names technology readiness, named industries, or a technical review panel is underwriting science. A programme that says 'any industry' and asks for a pitch deck is underwriting a team and a market.
- Advanced Industries Accelerator Grants (Colorado OEDIT) is built around named advanced industries and funds up to $250,000 depending on grant type, with technical milestones written into the award, not just a market opportunity.
- NLnet Foundation Grants funds specific technical work, not a company narrative, in amounts between €5,000 and €50,000 for a first grant, with the next deadline on November 3rd 2026, 12:00 CET.
- Ben Franklin Technology Partners funds $100,000 to $500,000 specifically as a regional technology-commercialisation mandate, which is a different review bar than a generalist seed fund.
- By contrast, Techstars at $120,000 and EWOR Fellowship at up to EUR 300k (Ideation) or EUR 500k (Traction) will take a deep tech company, but their diligence is built to evaluate a founder and a market, not a technical claim, so you need the science proven enough to pitch in five minutes before you apply.
The three shapes the money actually takes
Deep tech capital at this stage comes in three forms, and each one charges a different price. Knowing which one you're applying to before you write the application changes what you should put in it.
- Direct equity investment. Virginia Venture Partners (CIT GAP Funds) puts up to $125,000 as an initial investment, and Missouri Technology Corporation IDEA Fund invests via equity or convertible debt at amounts that vary by deal. You give up ownership, but spending is generally unrestricted once the technical case is accepted.
- Non-dilutive grants with reporting obligations. Rhode Island Innovation Voucher offers up to $75,000, non-dilutive, and Launch Minnesota Innovation Grants offers up to $50,000. You keep all the equity, but the money is usually tied to specific deliverables and you report against them, which means the grant shapes your sequence of work whether or not that sequence is the fastest path to proving the science.
- Matching or cost-share grants. Florida High Tech Corridor Matching Grants goes up to $150,000, matched, and some Advanced Industries Accelerator Grants tracks also require matching funds. No equity is taken, but you must bring cash or an industry partner's cash to the table dollar for dollar, which is the hardest condition on this list for a pre-revenue science company to meet.
The condition that disqualifies most applicants in this sector
It isn't the technical bar that eliminates most deep tech applicants. It's the matching requirement. A programme that asks you to match $150,000, like Florida High Tech Corridor, or to match funds under some Advanced Industries Accelerator Grants tracks, is assuming you already have cash on hand or an investor willing to co-sign on an unproven technology. A company whose product doesn't exist yet usually has neither. There's no revenue to draw the match from, and investors are reluctant to put money against a milestone that might fail for purely scientific reasons. This is why a researcher reading a programme's eligibility page should go straight to the word 'matched' or 'cost-share' before reading anything else. If it's there, ask honestly whether you can actually produce your half before you spend a week on the application.
If you can only write one application this month
Apply to NLnet Foundation Grants first. It funds the technical work itself, not a company story, in amounts from €5,000 to €50,000 for a first grant, which is exactly what unfinished science needs: money attached to a specific deliverable rather than to a growth metric. There's no matching requirement, no state incorporation test, and a fixed next deadline of November 3rd 2026, 12:00 CET, so you know precisely what you're building toward. If you're incorporated in Rhode Island, the second application worth your time is the Rhode Island Innovation Voucher, non-dilutive up to $75,000, with the next round accepting applications from October 6. Save the matching grants and the equity investments for after you have one technical milestone cleared and documented. They'll ask for proof you don't have yet.
The full list of the 48 programmes we've checked for science and deep tech eligibility, including the ones not named here, is on the science and research funding page.