Phase I funds a feasibility study, not a product. The 2026 SBA guideline lets agencies award up to about $323,090 without special approval, but actual awards vary widely: NSF funds up to ~$305K over 6–18 months, NIH around the guideline (and above it for approved topics), DOE $200–250K, NASA $150–225K over 6 months, many DoD topics $100–250K, and some Air Force open topics have historically run smaller and faster. Expect 6–12 months of performance, and expect to do at least two-thirds of the work in-house.
What wins Phase I is credible evidence the physics works and a team that can execute — see how proposals are scored.
Phase II is the real R&D money: currently up to about $2,153,927 under the SBA guideline, typically over 24 months, to take the Phase I result to a working prototype. Agencies weigh your commercialization plan heavily here. Roughly 40% of Phase I awardees convert to Phase II. Several agencies stack more on top: NSF's Phase IIB supplement matches third-party investment, NIH funds Phase IIB continuations, NASA offers Phase II-E/II-X extensions, and DoD writes sequential Phase IIs.
Phase III involves no SBIR money at all. It's any follow-on work that "derives from, extends, or completes" your SBIR project, paid with regular procurement or private dollars. Two things make it valuable: agencies may award Phase III contracts to you sole-source, with no further competition, and your SBIR data rights protect your technical data for 20 years. In defense, that combination is why primes and program offices take SBIR-funded startups seriously. But it's a right, not a promise — you still need a customer with budget.
Phase I is a feasibility study on a short clock, and at many agencies the core deliverable is analysis: CFD, FEA, thermal, or electromagnetic modeling that shows the concept holds. Teams that already have simulation running on day one spend the award proving, not tooling up. Ask us what an Ansys evaluation could cover for your Phase I workload.
See if you qualify for an Ansys eval + engineering support at no cost How access works →| Agency path | Decide → first dollar (realistic) |
|---|---|
| DoD (fastest paths: AFWERX open topics) | ~5–9 months |
| NSF (Project Pitch → invite → award) | ~7–10 months |
| NIH (study-section review cycle) | ~9–14 months |
Plan cash accordingly, and start registrations at least eight weeks before any deadline — the full chain is in the registration checklist.
Official sources for this guide: SBIR.gov — program basics · SBIR.gov — data rights tutorials · NIH TABA notice NOT-OD-26-075. Figures change; always confirm on the official page before relying on them.
