NSF SBIR Phase I Winners and Award History

According to NSF's solicitation NSF 26-510 (opens in a new tab), posted May 22, 2026, Phase I provides up to $305,000 for projects of 6 to 18 months and Phase II provides up to $1,250,000. According to the NSF SBIR/STTR program page (opens in a new tab), the Phase II amount covers 24 months. According to SBIR.gov (opens in a new tab), agencies may issue a Phase I award up to $335,937 and a Phase II award up to $2,239,578 without SBA approval as of October 16, 2026. NSF's $305,000 sits below that SBA guideline, so the NSF number is the one that limits your budget.

According to the NSF Seed Fund Phase I active awardee page, 111 companies held active Phase I awards when Grantsights checked it on October 18, 2026, with an average award of $302,003 across 37 states and territories. According to the NSF Seed Fund award search page (opens in a new tab), its records cover startup and small-business awards since 2013 and point older records to NSF Award Search (opens in a new tab).

NSF's Seed Fund pages list who won and how much, but a founder still has to decide whether the company's technical risk is ready for a pitch now, needs another round of lab proof, belongs at NIH or DOE instead, or isn't an SBIR project at all. The award patterns below are meant to help with that call.

10 NSF Phase I Winners From The June To August 2026 Restart

All 86 FY2026 Phase I awards landed in a 12-week window: 40 in June, 13 in July and 33 in August 2026. June 17, 2026 alone carried 20 awards. Here are 10 of them, drawn from the NSF Awards API and listed on the Seed Fund active awardee page.

For example, Lepidext received $305,000 in Kentucky, a state outside the top five, for a biopesticide discovery project that NSF also coded to its EPSCoR program element. One company in Puerto Rico, Cultimar Technologies, won $304,645 for aquaculture protein filtration. Those two awards show that NSF funds hard agricultural and biological problems from places that rarely appear in startup press coverage.

Caleo Biotechnologies' $289,435 award was the lowest of the 86. Every other FY2026 award was $300,000 or more, which tells you NSF expects a full-scale Phase I work plan, not a trimmed one.

Next route: Use the NSF SBIR guide, SBIR Phase I index, SBIR complete guide, SBIR first-time applicant guide, how to write an SBIR proposal, and the NSF grant listings before writing.

8 Recent 2025 NSF Phase I Winner Records From Award Search

FY2025 produced 195 Phase I awards: 145 SBIR and 50 STTR. The eight below come from the same NSF Awards API pull. Seven are SBIR Phase I awards; award 2528301 (Clarus Labs) is titled STTR Phase I in NSF Award Search (opens in a new tab).

None of these are payroll grants, launch grants, or ordinary software builds. Each title names a technical problem that might not work, such as a hydrogel that has to hold under bleeding conditions or a fermentation process that has to run hot. That's the pattern NSF funds.

Reef Arches' $154,646 award was the smallest FY2025 Phase I award in the data, and only 3 of 195 came in under $250,000. Asking for much less than the cap isn't a safer bet. It usually signals that the team hasn't scoped enough technical work to fill 6 to 18 months.

3-Year NSF Phase I Pattern From FY2024 To FY2026

Grantsights analysis of 584 NSF SBIR and STTR Phase I awards across FY2024, FY2025 and FY2026 shows award counts falling by more than two-thirds while the typical award grew to $305,000.

Three things stand out. First, the FY2025 drop came before the lapse: NSF made 195 awards that year, down from 303, and 84 of those 195 (43%) were dated July through September 2025. Second, the FY2026 count of 86 reflects only three months of awards, because NSF couldn't make new SBIR or STTR awards from October 1, 2025 until reauthorization. Third, STTR rose from 15.5% of Phase I awards in FY2024 to 29.1% in FY2026.

Don't read the record counts as a success rate. NSF Award Search shows awards, not unfunded applications. Project Pitch invitations, program director fit, company readiness and topic alignment all sit outside these numbers.

5 Top State Signals In The 2025 Pull

California led all three years, but its share fell from 26.7% in FY2024 to 19.5% in FY2025 and 18.6% in FY2026.

California and Massachusetts together held 38.3% of FY2024 Phase I awards and 27.9% of FY2026 awards. In FY2026, North Carolina and New Jersey each had 5 awards, and Kentucky had 3, more than Colorado or New York.

State concentration doesn't mean companies elsewhere should skip. It means you should compare yourself to the right peer set. A rural advanced-manufacturing company shouldn't treat a California biotech pattern as its benchmark. For founders in EPSCoR jurisdictions, the 12 FY2026 awards co-coded to NSF's EPSCoR program element are a real signal: NSF puts extra money behind strong projects from states with less research funding.

24 Topic Areas: Where FY2026 Phase I Awards Went

According to NSF's Seed Fund Phase I active awardee page, the 86 FY2026 awards fall into 20 of NSF's 24 listed technology topic areas. Health-related topics took 28 of 86 awards (32.6%).

The FY2025 pattern looks similar in NSF's own program codes. Of 195 FY2025 awards, 28 (14.4%) were coded Biomedical Engineering and Diagnostics, 16 (8.2%) Advanced Manufacturing, and 9 (4.6%) Artificial Intelligence and Cognitive Science.

The honest read for software founders: Artificial Intelligence was only 2 of 86 FY2026 awards. AI still appears in more than 10 of the 86 FY2026 award titles, from Channel Robotics' AI-enhanced endoscopic instruments to Graded Gaming's AI teacher, but those awards went to a hard technical problem in a domain, not to AI as the product. If your pitch is an AI wrapper on an existing model, NSF isn't the best route.

What Changed In 2026: Lapse, Reauthorization, And 4 New Deadlines

SBIR and STTR authority lapsed on October 1, 2025. According to the Congressional Research Service (CRS Insight IN12705 (opens in a new tab), July 13, 2026), the Small Business Innovation and Economic Security Act (P.L. 119-83), signed April 13, 2026, ended a six-month lapse and extended both programs through September 30, 2031.

According to NSF's restart announcement (opens in a new tab) of May 26, 2026, NSF relaunched SBIR/STTR with $250 million, including a $40 million pilot for next-generation scientific instrumentation under a separate solicitation, NSF 26-511. According to the Seed Fund critical information page (opens in a new tab), NSF began accepting new Project Pitches on June 2, 2026.

Deadlines are 5 p.m. in the submitting organization's local time. The Phase I amount didn't change with reauthorization, and the FY2026 award data confirms it: 85 of 86 awards were $300,000 or more.

4 Winner Lessons For NSF SBIR Phase I Applicants

The first lesson is that Phase I is a technical-risk award worth up to $305,000. NSF 26-510 centers R&D, proof-of-concept work, and commercial potential. A proposal that reads like normal product development fails because there's no research question for reviewers to score.

The second lesson is that the Project Pitch is a real gate. According to NSF 26-510, proposers must hold an official invitation, invitations are valid for the next two submission deadlines, and a company can submit at most two Project Pitches in a 12-month period, with no more than 3 pitches for the same technology. According to the NSF Project Pitch page (opens in a new tab), the pitch has four parts: the technology innovation and the technical objectives and challenges (up to 3,500 characters each), and the market opportunity and the company and team (up to 1,750 characters each). The page also says a company can have only one pitch pending at a time.

The third lesson is timing. The Project Pitch page doesn't state a response time. If you haven't pitched yet, the November 4, 2026 deadline is 17 days from this update, so plan for March 4, 2027 instead and use the extra months to firm up data. A common mistake is pitching a half-formed idea to "hold a spot": a rejected pitch uses one of your two pitches for the year.

The fourth lesson is that SBIR and STTR follow different partner rules. STTR requires a formal research-institution role; SBIR doesn't. STTR's share of NSF Phase I awards rose to 29.1% in FY2026, so a strong university partnership is clearly fundable, but a decorative partner is a red flag. Compare the rules in the SBIR vs STTR differences guide before you choose.

What NSF SBIR Phase I Winners Don't Tell You

Winner lists cover 584 funded awards over three fiscal years, but they don't tell you whether NSF will invite your Project Pitch.

Winner lists also hide timing. A founder who looked at FY2026 awards in May 2026 would have seen zero, because the lapse froze new awards for eight months of the fiscal year. Anyone who read that as "NSF stopped funding startups" would have missed a $250 million restart. Check the award date range before you draw conclusions from any winner list.

Finally, the lists don't show how many companies tried. Founders often underestimate how much work the pitch-to-proposal path takes: a Project Pitch, an invitation, a full proposal with a commercialization plan, then review. Read the SBIR commercialization plan guide before you commit.

6 Apply, Wait, Or Skip Checks

Use these 6 checks before a founder spends weeks on the Project Pitch and full proposal.

Skip NSF if your main risk is sales, not science. NSF won't fund a company to finish a product that already works. Biomedical teams should compare institute fit in the NIH SBIR/STTR guide, and teams with a defense or energy customer should read the SBIR agencies comparison. If you already hold a Phase I, the SBIR Phase 2 guide covers the $1,250,000 next step.

How To Run Your Own NSF Award Search In 4 Steps

These 4 steps take about 30 minutes and give you a peer set for your Project Pitch.

1. Open NSF Award Search (opens in a new tab) or the NSF Seed Fund award search (opens in a new tab) and filter to Phase I.
2. Narrow to your topic area and the last two or three fiscal years. Remember FY2026 awards only run June through August 2026.
3. Read the abstracts of the 5 to 10 closest winners. Note how each one states the technical risk in its first paragraph.
4. Compare their budgets to yours. If almost every peer award sits between $300,000 and $305,000, a much smaller request needs a reason.

For the cross-agency view, see the SBIR Phase I winners guide. Then browse the free grant database to compare NSF SBIR Phase I with other programs before writing.

Sources

• NSF SBIR/STTR program page - seedfund.nsf.gov (opens in a new tab)
• NSF SBIR/STTR solicitation NSF 26-510 - nsf.gov (opens in a new tab)
• NSF restart announcement, May 26, 2026 - nsf.gov (opens in a new tab)
• NSF Seed Fund critical information - seedfund.nsf.gov (opens in a new tab)
• NSF Project Pitch information - seedfund.nsf.gov (opens in a new tab)
• NSF Seed Fund Phase I active awardees - seedfund.nsf.gov (opens in a new tab)
• NSF Seed Fund award search - seedfund.nsf.gov (opens in a new tab)
• NSF Award Search - nsf.gov (opens in a new tab)
• NSF Awards API - api.nsf.gov (opens in a new tab)
• SBIR.gov award size guidelines - sbir.gov (opens in a new tab)
• SBIR.gov data resources - sbir.gov (opens in a new tab)
• Congressional Research Service, CRS Insight IN12705 - congress.gov (opens in a new tab)

Related Guides

• NSF SBIR guide
• SBIR grants complete guide
• SBIR first-time applicant guide
• How to write an SBIR proposal
• SBIR agencies comparison
• SBIR vs STTR differences
• SBIR Phase I index
• SBIR Phase I winners across agencies