If you’re building in deep tech — hardware, novel materials, AI infrastructure, biotech, energy systems, or anything with real R&D risk — SBIR grants are probably the single largest pool of non-dilutive capital you qualify for and haven’t touched yet. The Small Business Innovation Research program moves more than $4 billion a year across 11 federal agencies, and most eligible pre-seed founders never apply, usually because the process looks more intimidating than it actually is.
Here’s what’s changed for 2026, which agencies are worth your time first, and how to approach your first Phase I application without burning your entire runway on paperwork.
Why SBIR Matters More in 2026
As of April 2026, agencies can issue Phase I awards up to roughly $323,000 and Phase II awards up to about $2.15 million without needing special SBA approval — a meaningful step up from prior thresholds. That means the ceiling on non-dilutive capital available to a deep tech startup with zero product revenue is higher than most founders realize. Combined with rising federal interest in reshoring critical technologies (semiconductors, energy storage, advanced manufacturing, defense-relevant AI), 2026 is a strong year to be a technical founder chasing “SBIR grants 2026” rather than a Series A term sheet.
The Top SBIR Agencies for Deep Tech Founders
Not all 11 participating agencies are equally relevant to a deep tech startup. Four consistently deploy the most capital and run the topics most aligned with hard-tech founders.
1. Department of Defense (DOD)
DOD is the largest SBIR spender by award volume and the best fit for founders working on autonomy, robotics, sensing, cybersecurity, advanced materials, or dual-use hardware. Typical Phase I awards run in the $150,000–$250,000 range over roughly 6 months, with fast-track and direct-to-Phase-II pathways increasingly common for startups that can show early technical traction. DOD’s component agencies (Army, Navy, Air Force, DARPA, SOCOM, and others) each run their own topic calls multiple times a year, so a single “no” from one component doesn’t mean the technology is a poor fit elsewhere.
2. National Institutes of Health (NIH)
NIH is the agency of choice for biotech, medtech, diagnostics, and health-adjacent deep tech. Standard Phase I budgets sit around $300,000, and NIH is notably founder-friendly about funding genuinely early-stage science — you don’t need existing revenue or a finished prototype, just a credible research plan and a qualified team. NIH also runs rolling submission cycles rather than a single annual deadline, which gives founders more flexibility to time an application around their runway.
3. Department of Energy (DOE)
DOE is the natural home for founders in energy storage, grid technology, advanced manufacturing, nuclear, or materials science. Award sizes are competitive with DOD and NIH, and DOE topics are often written with unusual technical specificity — a strong signal for founders who can show they understand the underlying physics or process chemistry better than a generalist reviewer would expect. DOE’s SBIR office also has an active track record of funding startups through Phase II and into pilot deployments with national labs.
4. National Science Foundation (NSF)
NSF SBIR runs a flat $305,000 Phase I award and is arguably the most startup-native of the major agencies — its process was explicitly designed for commercially-oriented small businesses rather than academic labs. NSF is a strong fit for founders in AI/ML infrastructure, novel software-hardware systems, and general-purpose deep tech that doesn’t map neatly onto a single mission agency. NSF Phase II can reach roughly $1.25 million over 24 months, and the agency is known for a genuinely fast, founder-readable review process.
Eligibility: The Basics That Trip Founders Up
Across all four agencies, the baseline eligibility rules are consistent:
- For-profit, U.S.-based small business — more than 50% owned by U.S. citizens or permanent residents, and no more than 500 employees.
- The principal investigator’s primary employment must be with the small business at the time of award (a common surprise for academic-founder teams still splitting time with a university).
- No product on the market yet is not disqualifying — SBIR is explicitly built for pre-revenue, pre-prototype R&D risk. This is why it’s such a strong fit for pre-seed deep tech founders who don’t yet have traction a VC would fund.
- STTR vs. SBIR — if your technology originated in a university lab and you need a formal research partnership with that institution, look at the STTR variant instead, which requires a subcontract to a research institution.
How Pre-Seed Founders Should Approach Phase I
Start with fit, not with writing. Before you draft a single sentence, map your technology against actual open topic calls at DOD, NIH, DOE, and NSF. A technically strong proposal aimed at the wrong topic loses to a modest proposal aimed at exactly what the program officer is looking for.
Treat Phase I as a 6-month feasibility study, not your whole roadmap. Reviewers are funding a bounded, answerable technical question — not your five-year vision. Scope the proposal to what you can credibly prove in the award period with the budget you’re requesting.
Call the program officer before you submit. This is the single most underused move in SBIR strategy. Program officers want technically strong applications in their topic area and will often tell you directly whether your approach fits before you invest weeks writing it.
Budget your time, not just your dollars. A competitive SBIR proposal takes real hours — typically 40–80 across narrative, budget justification, and letters of support — and that time is a direct trade against the customer conversations and product work that build your traction story in parallel. This is exactly the tension we built the SuperConnector Club’s methodology around: balancing funding-track effort against traction-track effort so neither one starves the other.
Apply in parallel, not sequentially. Nothing stops you from submitting to DOD and NSF in the same quarter if your technology genuinely fits both. Federal agencies run independent review cycles, and stacking applications is how founders assemble a real non-dilutive war chest instead of waiting on a single yes/no.
Where to Go Next
If you’re mapping out which non-dilutive sources you actually qualify for, our Funding Hub keeps a live, curated list of grant programs, competitions, and revenue-based capital options beyond SBIR alone. And if SBIR specifically is your next move, our free SBIR & STTR founder’s playbook walks through topic selection, budget structuring, and proposal narrative in more depth than any single blog post can.
The Takeaway
SBIR isn’t a consolation prize for founders who can’t raise venture capital — for deep tech companies with real R&D risk, it’s often the smartest capital on the table: non-dilutive, sized for genuine technical work, and available well before you have the traction a VC would require. Start by matching your technology to the right agency, treat Phase I as a scoped feasibility bet, and build your application timeline into your broader funding-versus-traction balance rather than treating it as a side project.