Technology and Software Business Grants 2026

The technology funding landscape in 2026

349 programs touch technology; 218 of them are open right now

The GrantCompass US catalog holds 736 programs, and 349 of them touch technology or software — 346 by the industry field itself, plus a handful reached only through program tags. That is second in the catalog only to manufacturing, which 373 programs serve by the same field. Of those 349, 218 are open or rolling as of August 28, 2026; the remaining 131 sit between intakes, run on annual cycles, are upcoming, or have ended, and several of them are the best programs in the set. 87 of the open 218 are grants in the strict sense: money that is neither repaid nor delivered through a tax return. The 131 open programs that are not grants are 57 tax credits, 43 services and advisory programs, 26 loans, 4 competition awards and 1 forgivable loan. The median published ceiling across the technology slice is $250,000, against $150,000 catalog-wide — technology programs run larger than US small-business funding as a whole. 227 of the 349 publish a ceiling at all; the rest are negotiated, percentage-based, or uncapped.

Technology funding concentrates into three durable layers

The 349 programs are not 349 separate ideas. US technology funding concentrates into three layers that stack on one another, and a company doing genuine research can hold all three in the same tax year.

• Federal R&D grants (SBIR and STTR) -- 11 federal agencies are required by law to set aside a portion of their external research budgets for small businesses. This is the largest source of non-dilutive funding for technology startups in the US, spanning software, AI, hardware, biotech, space, energy, defense, and ed-tech. Awards run from roughly $150,000 to over $2 million depending on agency and phase.
• The federal R&D tax credit (IRC Section 41) -- a permanent, non-competitive credit that rewards spending on qualified research. Software development is one of the most common qualifying activities. Unprofitable startups can take the credit against payroll taxes, turning R&D spend into cash even with no income.
• State programs -- two stackable sub-layers: state SBIR/STTR matching grants that add money on top of a federal award, and state R&D tax credits that sit on top of the federal credit. Both are low-competition relative to the federal grants themselves.

A fourth category -- accelerators, proof-of-concept grants, revenue-based financing, and SBA-backed loans -- fills the gaps for companies that aren't ready for SBIR or aren't doing novel research. The largest federal programs of all (Department of Energy, EDA, and NSF consortium awards in the tens or hundreds of millions of dollars) exist, but they fund consortia and deep-tech scale-ups, not typical small businesses; they are context, not an action item, for most readers.

The same research spending can be funded twice

Six instruments carry almost all US technology funding

Six instruments carry nearly all of the money in the technology slice, and they differ far more in what they may pay for than in size. A grant pays for work described in advance in a proposal. A tax credit pays back money already spent, after the fact, through a return. A loan is repaid. A state SBIR match is close to automatic once a federal award lands. Reading a program's instrument before its dollar figure is the single fastest way to rule it in or out.

Most "technology grants" are research grants — and that decides whether they are for you

The largest technology funding programs in the United States are research programs. SBIR and STTR — the federal awards that dominate every list of "tech startup grants" — pay a small business to perform research and development that removes a specific technical uncertainty. They do not pay to build a product, hire a sales team, acquire customers, or extend a working SaaS platform. NSF states the boundary plainly on its America's Seed Fund site: "We support research and development of deep technologies - those that are based on discoveries in fundamental science and engineering. As we review applications, we consider your technology's innovativeness, commercial potential, and possible societal impact" (seedfund.nsf.gov, read August 28, 2026). Commercial potential is a scoring criterion there; commercial work is not the funded activity. Almost no competing page draws that line, and the cost of not drawing it is a wasted month writing a proposal that was never eligible.

What a research grant pays for, and what it does not

A federal research award pays the direct cost of performing a defined research task: the salaried hours of the technical staff doing the work, materials and test articles consumed by it, subcontracted research, and a negotiated indirect-cost rate on top. NSF SBIR Phase I funds that work up to $305,000; NIH SBIR Phase I up to $323,090; Phase II runs to $1,000,000 at NSF and $2,153,927 at NIH. The same award will not fund the ordinary business around the research — sales and marketing, customer acquisition, routine feature work, general working capital, or buying an existing product. The working test is three steps. First, name the technical question that currently has no known answer. Second, describe the experiment that would answer it. Third, state what a negative result would look like. A project that cannot survive the third step is product development, and belongs in a different funding conversation.

Where the line falls for a software company

Software is fundable when the difficulty is technical rather than commercial. A new algorithm whose accuracy or latency is genuinely unknown; a distributed system whose behaviour at a scale nobody has run is uncertain; a model whose performance on a hard domain has not been established; a protocol or cryptographic scheme that needs proving — these are research questions, and federal agencies fund them. Building a well-understood web application faster, adding an integration, redesigning a workflow, or porting an existing product to a new platform is engineering: difficult, valuable, and not research. The federal R&D tax credit under IRC Section 41 applies a related but looser standard, which is why a software company that would never win an SBIR award can still claim the credit — Section 41 asks for technical uncertainty resolved through a process of experimentation, not for novelty measured against the state of the art.

Only 14 of the 87 open technology grants name software

The distance between "technology" and "software" is measurable in the GrantCompass catalog itself. Of the 218 technology programs open as of August 28, 2026, 87 are grants — and 14 of those 87 list software among the industries they serve. Across all 218 open technology programs of every instrument, 46 name software. The industries that co-occur with technology tell the same story: 196 of the 218 also serve manufacturing, 113 serve healthcare, 111 agriculture, 105 clean technology, 91 life sciences and 78 aerospace. A technology program in the United States is, statistically, a program for something physical — a device, a material, a molecule, a machine, a plant — that a technology company happens to be able to reach. The single largest open grant that names software is NIH SBIR Phase II at $2,153,927, which is a health research award.

If your company is not doing research, start here instead

A technology company that performs no research is not shut out of public money; it is shut out of the grant layer, which is a far smaller claim. Four pots stay fully open. The federal R&D tax credit and 32 active state R&D credits pay back qualifying development spending on a standard looser than SBIR's. 27 state workforce programs pay to train the engineers already on the payroll. 22 state job-creation credits pay per engineer hired. Six state and federal export grants pay to sell a technology product abroad. Those groups — with equipment money, accessible debt and investor-side credits — are inventoried further down this page under what the money is actually for. None of them is badged as a technology program, which is exactly why technology companies leave them unclaimed.

Which technology companies are served generously, and which are served thinly

US technology funding is not spread evenly across technology. It follows federal missions — health, defence, energy, agriculture, standards, water, transport — so a company whose work advances one of those missions has many more doors than a company whose work advances a customer's revenue. Counting the 218 technology programs open as of August 28, 2026 by the other industries each one serves produces a clear ranking: healthcare 113, clean technology 105, life sciences 91, aerospace 78, supply chain 63, biotechnology 57, software 46, fintech 40. Deep tech, hardware, life-science tools and dual-use defence are served generously and by several agencies each. A business-to-business SaaS company with no research component is served by a fraction of the same catalog, and mostly through tax credits and general small-business programs rather than grants. Each slice below names the programs that actually serve it.

Health technology and life sciences are the best-served slice

Health is the largest single destination for US technology money. 113 of the 218 open technology programs also serve healthcare, 91 serve life sciences and 57 serve biotechnology, and the awards are the largest in the set. NIH SBIR Phase I pays up to $323,090 and Phase II up to $2,153,927 — the full SBA ceilings, which NIH uses while most agencies award below them. ARPA-H Innovative Solution Openings run from roughly $1 million to $50 million on Other Transaction agreements rather than grants, and accept submissions year-round. BARDA DRIVe funds health-security ventures on a project basis. A digital-health, clinical-decision or medical-device company should treat NIH as its default agency rather than NSF. Deep guide: healthcare and biotech business grants.

Deep tech, hardware and instruments: NSF, NIST and the national labs

Deep-tech hardware is what the phrase "technology grant" was built for. NSF's America's Seed Fund funds research and development of deep technologies grounded in fundamental science and engineering, up to $305,000 at Phase I and $1,000,000 at Phase II, entered through a 3,500-character Project Pitch. Outside SBIR, NIST's Measurement Science and Engineering research grants fund metrology and standards work, the LPS Qubit Collaboratory funds quantum-information research at up to $800,000 a year, and the Naval Research Laboratory's Long Range Broad Agency Announcement accepts basic and applied research proposals continuously. Hardware companies also reach equipment and adoption money that pure-software companies cannot use, which is inventoried later on this page.

Defence and dual-use is where pure software is funded generously

Defence is the one part of the US system that funds software as software, at scale, and on a fast cycle. The Air Force's AFWERX Open Topics accept proposals continuously rather than on fixed cycles and fund dual-use software up to $250,000 at Phase I; Department of Defense SBIR Phase II runs to $2,000,000. OSD ManTech's Advanced Manufacturing Technology program funds $150,000 to $8,800,000 for manufacturing-technology work. The DoD Mentor-Protégé Program subsidises a large prime contractor to develop a small supplier. Defence agencies issue SBIR as FAR-based contracts with deliverables and milestones rather than as research grants, which suits a company that wants a government customer and frustrates one that wants latitude.

Energy, climate and grid technology

Energy is the second-largest destination: 105 of the 218 open technology programs also serve clean technology. The DOE Office of Energy Efficiency and Renewable Energy keeps multiple Funding Opportunity Announcements open at any time, at $500,000 to $20 million or more per award with cost-share required. The DOE Office of Science FY2026 solicitation funds $50,000 to $5,000,000 a year of basic research including computing and physics. NREL's American-Made Challenges run prize competitions from $50,000 to $3 million plus national-lab vouchers, which is a far more accessible on-ramp for a small company than a consortium award. DOE's advanced nuclear licensing cost-share covers regulatory work. Deep guide: cleantech business grants.

Semiconductors and advanced manufacturing technology

Semiconductor and production-technology work is funded through its own channels rather than through SBIR. The NIST CHIPS R&D Office Broad Agency Announcement has a $10 million minimum and no stated ceiling — a consortium instrument, not a startup one. The Texas Semiconductor Innovation Fund negotiates awards in the millions. Manufacturing USA institute project calls, including America Makes, run $1 million to $12 million or more per call and are usually team-based. At the small end, Michigan's Industry 4.0 Technology Implementation Grant pays up to $25,000 at a 50% match and Connecticut's Additive Manufacturing Voucher up to $20,000. Deep guide: manufacturing business grants.

Agriculture technology, water and environmental sensing

Agricultural and environmental technology is served by mission agencies with modest ceilings and predictable annual cycles. USDA NIFA's SBIR Phase I funds up to $175,000 — the lowest Phase I ceiling among the major agencies — and releases one solicitation a year. NOAA's National Bycatch Reduction Engineering Program funds $50,000 to $400,000 for gear and sensing engineering. The Bureau of Reclamation's Desalination and Water Purification Research program funds $100,000 to $800,000 for water-treatment technology, including pilot-scale testing. Agricultural technology is also the one slice where the SBIR ceiling is materially lower than elsewhere, so a company able to frame its work for either USDA or NSF should compare the two before choosing. 111 of the 218 open technology programs also serve agriculture, more than serve clean technology. Deep guide: farming and agriculture grants.

B2B SaaS and IT services are the thin end of the market

A business-to-business SaaS company with no research component is the least-served technology business in the United States, and pretending otherwise wastes its time. 46 of the 218 open technology programs name software; 14 of those are grants, and nearly all of the 14 are mission research awards that a general SaaS product cannot address. There is exactly one state program in the GrantCompass catalog written for software companies as such: Louisiana's Digital Interactive Media and Software Program, which pays 25% of qualified in-state payroll for digital interactive media and software development. Everything else that works for a SaaS or IT-services firm is either a tax credit on development spending, workforce and hiring money, accessible debt, or free advisory support — all of it inventoried below, none of it labelled "technology grant".

SBIR and STTR are the core non-dilutive funding for tech startups

The Small Business Innovation Research (SBIR) program and its sibling the Small Business Technology Transfer (STTR) program are the foundation of US technology grant funding. By statute, 11 federal agencies set aside a percentage of their extramural R&D budgets to fund research and development performed by US small businesses. SBIR is the single largest source of non-dilutive early-stage capital for technology companies in the country. STTR is the same idea with one added requirement: the company must partner with a research institution (a university or federal lab) that performs at least 30% of the work.

The US Small Business Administration describes the pair in one line on sbir.gov: "SBIR provides equity free funding through federal agencies to American small businesses" (sbir.gov/about, read August 28, 2026). This page covers what SBIR and STTR are, what they will and will not fund, and which agencies suit which technology. It deliberately does not repeat the mechanics — phase-by-phase eligibility, budget construction, the reviewer's scoring frame, and the full 11-agency comparison are the subject of the dedicated guide, SBIR grants for startups. Which SBIR and STTR windows are accepting proposals today is answered on SBIR and STTR grants open now.

How the phases work

SBIR and STTR money arrives in stages, and which stage a company is eligible for is determined entirely by what it has already won, not by how far along the technology is. A first-time applicant can only enter at Phase I, regardless of maturity, and Phase II is closed to any company that has not completed a Phase I on the same technology at the same agency.

Both programs are structured in phases. Phase I funds feasibility -- proving the core technical concept works -- and is the entry point. Phase II funds full development of a prototype or product and is open only to companies that completed Phase I. (A later Phase III is the commercialization stage, funded by non-SBIR dollars such as government contracts or private capital.) Eligibility for both phases is the same: a US-owned, for-profit business with 500 or fewer employees, with the principal investigator primarily employed by the company at award.

Award ceilings by agency

Phase I and Phase II award ceilings are set by SBA and adjusted for inflation. As of April 2026, agencies may award up to $323,090 at Phase I and $2,153,927 at Phase II without an SBA waiver; individual agencies often set their own amounts below those ceilings (NSF, for example, funds Phase I up to $305,000). The 2026 reauthorization also added a Phase II "Strategic Breakthrough" track that lets the largest agencies award up to $30 million to a single firm.

SBA publishes the governing numbers directly: "As of April 2026, agencies may issue a Phase I award (including modifications) up to $323,090 and a Phase II award (including modifications) up to $2,153,927 without seeking SBA approval. Any award above those levels will require a waiver" (sbir.gov/about, read August 28, 2026). Those are ceilings, not typical awards — NSF funds Phase I at up to $305,000 and USDA NIFA at up to $175,000 by their own choice.

The NSF Project Pitch is the usual front door

SBIR and STTR are authorized through September 30, 2031

The SBIR and STTR programs run on congressional authorization rather than permanently, and that authorization lapsed on September 30, 2025 before being restored. The programs were reauthorized through September 30, 2031 when the Small Business Innovation and Economic Security Act was signed into law in April 2026. As of mid-2026 the programs are authorized and agencies are issuing solicitations on normal cycles again, so you can safely build a timeline around them. NSF reopened Project Pitch submissions on June 2, 2026 in response to the new solicitations. The practical consequence for a founder is that a 2026–2027 funding timeline built on SBIR is safe to plan against, which was not true during the lapse.

Choosing between the SBIR agencies

Contract agencies behave differently from grant agencies

Two structural differences between SBIR agencies matter more to a technology company's day-to-day than the award ceiling does: whether the agency issues the award as a contract or a grant, and whether it publishes topics or accepts investigator-initiated ideas.

The 13 SBIR and STTR solicitations in this catalog

This page links 13 of the 19 SBIR and STTR solicitation records in the GrantCompass catalog — the agencies that matter most to technology companies — with published Phase I ceilings from $175,000 (USDA NIFA) to $323,090 (NIH) and Phase II ceilings to $2,153,927 (NIH). Ten of the 13 are active; the Department of Education's IES program and USDA NIFA sit between intakes on annual solicitation cycles, and the Department of Defense's Phase I is upcoming on its next fixed cycle. None of them takes equity and none is repaid. The catalog also carries additional agency solicitations — DARPA, NASA, Missile Defense Agency, Homeland Security — listed on the SBIR guide rather than repeated here.

Federal research money outside SBIR

SBIR is not the only way to be paid to do research by the US government, and the non-SBIR route is where the larger awards are. GrantCompass tracks 13 open federal research programs outside SBIR that serve technology companies, with published ceilings from $400,000 (NOAA's Bycatch Reduction Engineering Program) to $50,000,000 (ARPA-H Innovative Solution Openings) and a median published ceiling of $6,900,000. Most are Broad Agency Announcements or rolling solicitations rather than dated competitions, so a company can approach them when its work is ready. They are more procedural than SBIR: cost-share and a SAM.gov Unique Entity ID are common requirements.

Prize competitions and consortium calls

Prize competitions are the most accessible federal research money for a small technology company, because they are judged on a submission rather than on institutional capacity. NREL's American-Made Program prize challenges run from $50,000 to $3 million or more per challenge and add national-lab vouchers that buy access to instruments a startup cannot own. Manufacturing USA institute project calls run $1 million to $12 million or more and are usually consortium-led, so a small company typically joins as a partner rather than as prime. The SBA Growth Accelerator Fund Competition awards $75,000 to $150,000 — to the accelerators and incubators that support technology founders, not to the founders themselves, which is a distinction worth knowing before applying.

The federal R&D tax credit is the most universal technology-company benefit

If SBIR is the most valuable competitive program, the federal Research & Development tax credit under IRC Section 41 is the most universal one. It is permanent, non-competitive, and available to essentially any company doing qualifying technical work -- including software development. Where SBIR requires winning a competition, the R&D credit simply requires that you spent money on qualified research and can document it.

Three properties make the Section 41 credit unlike every grant on this page. It is permanent, so it can be planned around rather than chased. It is non-competitive: no reviewer decides whether a company deserves it, only whether the work and the documentation meet the statutory test. And it is recurring, so a software company doing qualifying development generates a claim every single year rather than once. The trade-off is that it arrives through a tax return rather than as a cheque, which means it rewards bookkeeping discipline more than it rewards a good proposal.

Why software development qualifies

Software development is one of the most common qualifying activities under Internal Revenue Code Section 41, and the reason is structural: writing software that has never been written before necessarily involves resolving uncertainty by trying things. The test the Internal Revenue Service applies is about the process, not about the novelty of the result against the wider industry.

Qualified research under Section 41 must meet a four-part test: it must be technological in nature, aimed at a permitted purpose (new or improved functionality, performance, reliability, or quality), involve the elimination of technical uncertainty, and proceed through a process of experimentation. Routine software work doesn't qualify -- but developing new product features, novel algorithms, new architectures, performance improvements, or integrations that aren't off-the-shelf typically does. Qualified expenses include the wages of engineers and developers, the cost of contract research, and computing/cloud costs consumed in development.

The payroll-tax offset is the provision that matters to startups

How the payroll-tax election is actually made

The Internal Revenue Service sets out the mechanics precisely. The Inflation Reduction Act's provision 13902 raised the maximum payroll-tax research credit a qualified small business may elect from $250,000 to $500,000 for tax years beginning after December 31, 2022, and from the first quarter of 2023 "the payroll tax credit is first used to reduce the employer's share of social security tax up to $250,000 per quarter and any remaining credit reduces the employer's share of Medicare tax for the quarter" (IRS, "Qualified small business payroll tax credit for increasing research activities", read August 28, 2026). The election is made on Form 6765 attached to a timely filed income tax return; Form 8974 then computes how much may be used in each quarter, and any excess carries forward.

State R&D credits stack on top of the federal credit

Most states offer their own R&D tax credit that piggybacks on the federal one, and several are unusually generous for small companies. These stack: you can claim both the federal credit and your state's credit on the same qualifying expenses.

Seven states pay cash for research even when the company is losing money

Refundability is what separates a useful state R&D credit from a theoretical one, because a pre-revenue technology company owes little or no state income tax for a non-refundable credit to reduce. Seven of the state R&D credits open as of August 28, 2026 pay cash or partial cash rather than only reducing tax owed. Hawaii's Research Activities Credit is fully refundable at 20% of Hawaii qualified research expenses. Delaware's credit is refundable at 20% or 10% of incremental Delaware QRE. New York's Excelsior R&D credit is refundable at 50% of the federal Section 41 credit, capped at 6% of New York research spending. Arizona refunds 75% of the credit for companies under 150 employees, and Connecticut, Minnesota and Wisconsin are partially refundable.

Thirteen more states credit research against tax you will owe later

Thirteen further state R&D credits open today are non-refundable but carry forward, which makes them valuable to a technology company that expects to be profitable rather than to one that expects to be acquired. Louisiana is the highest headline rate at 30% for firms under 50 employees, tiered down to 10% and 5% for larger ones. Vermont credits 27% of the Vermont-apportioned federal Section 41 credit. Rhode Island runs 22.5% then 16.9%. California is 15% incremental with unlimited carryforward, which for a long-lived software company is worth more than a higher rate with a short window. Indiana is 15% or 10%, Nebraska 15%, Georgia 10% against payroll withholding, and Illinois 6.5%.

Five technology credits that are not shaped like an R&D credit

Five programs reward technology spending through a mechanism other than a percentage-of-research credit, and each one reaches a company the standard credit misses. Louisiana's Digital Interactive Media and Software Program pays 25% of qualified in-state payroll for software and digital-media development — the only state program in this catalog written for software companies as such, and it does not require the work to be novel. Washington's R&D sales and use tax exemption waives sales tax on machinery and equipment used in research rather than crediting the wages. Kentucky's Qualified Research Facility Tax Credit works the same way from the other end: it credits 5% of what it costs to build, remodel or equip a research facility in Kentucky, and explicitly excludes wages, contractor services and supplies — which is why it sits here rather than with the thirteen research-spend credits above. New Jersey's Technology Business Tax Certificate Transfer Program lets an unprofitable technology or biotechnology company sell unused net operating losses and R&D credits for cash, up to $20 million over the company's life. And the federal Section 41 credit itself sits underneath all of them.

Section 174 governs how research costs are deducted

Beyond the credit itself, the underlying tax treatment of R&D costs has been a moving target in recent years. How domestic software-development and research costs are deducted (immediate expensing vs. multi-year amortization under IRC Section 174) materially affects a tech company's cash taxes, and the rules changed with 2025 legislation. The One Big Beautiful Bill Act, signed July 4, 2025, restored immediate expensing of domestic R&D costs (including software development) starting with the 2025 tax year, and lets eligible small businesses recover amounts they were forced to capitalize in 2022–2024.

State SBIR matches and innovation grants are the low-competition layer

The least-known and lowest-competition technology grants in the country are run by the states. They come in two forms: matching grants that add money on top of a federal SBIR/STTR award, and innovation or proof-of-concept grants for early-stage companies that haven't won federal funding yet.

State technology programs are worth more attention than their size suggests for one reason: their applicant pools are small. A federal SBIR solicitation is read by companies in every state; a Vermont matching grant is read by companies in Vermont that hold a federal award, which is a very short list. GrantCompass tracks 15 state SBIR/STTR matching programs and 17 state commercialization and proof-of-concept programs serving technology companies, plus four state capital funds written specifically for technology firms. Every one of them is administered by a state economic-development or technology agency rather than by a federal one, so the rules, windows and definitions differ from state to state and must be read locally.

State SBIR/STTR matching grants

A state SBIR match adds non-dilutive state money on top of a federal award you've already won. Competition is minimal because eligibility is essentially gated by the federal win itself -- if you have an active SBIR or STTR award and operate in the state, you typically qualify. Vermont's Elevate Vermont match is explicitly non-competitive and first-come, first-served, paying up to $50,000. Wyoming's match pays up to $100,000 for Phase I and $200,000 for Phase II on a rolling basis. New York's NYSTAR Innovation Matching Grants pay up to $100,000 (Phase I) and $200,000 (Phase II). Indiana's FAST Program adds up to $75,000 on Phase II awards plus proposal coaching, and MassVentures START in Massachusetts pays $100,000 to $500,000 to commercialize Phase II deep tech.

All 15 state SBIR/STTR matching programs in the catalog

GrantCompass tracks 15 state SBIR/STTR matching programs across 15 states, with flat ceilings from $5,000 (Utah's proposal-preparation microgrant) to $300,000 (Tennessee's LaunchTN match, for Phase II) and a median of $100,000. Ten are active as of August 28, 2026 — Alabama, Indiana, Kentucky, Montana, New Mexico, Utah, Vermont, West Virginia, Wisconsin and Wyoming. Four sit between intakes on fixed rounds: Alaska, Hawaii, New York and Tennessee. North Carolina's is upcoming. Alabama's Innovation Grant is the largest single-award match at up to $250,000, and Hawaii's HSBIR match is expressed as a percentage of the federal award rather than a flat sum. Eligibility in almost every case is gated on holding a federal SBIR or STTR award and operating in the state.

The match is money you have already qualified for

The 15 matching programs, state by state

All 15 state SBIR/STTR matching programs appear below, including the six omitted from the table above. Read your own state's entry before submitting the federal proposal rather than after the award lands: several matches require registering with the state agency in advance, and a match generally cannot be claimed once the federal award has been fully spent.

State innovation and proof-of-concept grants

For pre-revenue companies that haven't won (or aren't pursuing) SBIR, many states run their own innovation grants. The Arizona Innovation Challenge awards up to $100,000 in non-dilutive funding across spring and fall cycles. Colorado's Advanced Industries Accelerator offers $150,000 proof-of-concept and $250,000 early-stage grants with no equity taken. The Missouri Technology Corporation awards up to $100,000 to validate commercial viability, South Carolina's SCRA offers $25,000 to $50,000 non-dilutive startup grants, and Ben Franklin Technology Partners in Pennsylvania makes $50,000 to $500,000 seed investments repaid from revenue rather than equity. New Jersey's Technology Business Tax Certificate Transfer Program is a category of its own: unprofitable tech and biotech companies can sell unused net operating losses and R&D credits for cash -- up to $20 million lifetime -- without giving up any equity. And in Minnesota, the Angel Tax Credit gives your investors a 25% refundable credit, making a priced round materially easier to close.

Seventeen state commercialization and proof-of-concept programs

Proof-of-concept and pre-seed programs are the state answer to the gap between a research result and a product, and they are the most useful state money for a technology company that has not won a federal award. GrantCompass tracks 17 such programs, with published ceilings from $25,000 (Maryland TEDCO's Rural Business Innovation Initiative and South Dakota's Proof of Concept) to $500,000 (MassVentures START and Ben Franklin Technology Partners in Pennsylvania), and a median of $100,000. Several are not grants at all in form: Connecticut Innovations' Proof-of-Concept Fund uses a $50,000–$100,000 convertible note, and TEDCO's Seed Fund takes equity. Read the instrument before the amount.

State technology capital funds lend where banks will not

Four states run capital vehicles aimed specifically at technology companies rather than at small businesses generally, and they fill the gap between a proof-of-concept grant and a priced round. The Texas Product Development and Small Business Incubator Fund lends up to $5,000,000 for product development and commercialization. MassDevelopment's Emerging Technology Fund lends up to $4,000,000 for technology facilities and equipment. Wisconsin's Technology Development Loan covers up to 20% of project cost. California IBank's Venture Capital Program, funded through the State Small Business Credit Initiative, co-invests state money alongside private venture funds. These are debt and equity, not grants: they are repaid or they dilute.

What the money is actually for: equipment, people, and premises

Grouping the 218 open technology programs by research area answers "who gets funded". Grouping them by what the money may be spent on answers the more useful question for a company with a budget. Outside research and its tax credits, the technology slice resolves into five purposes, and every one of them is available to a company that does no research at all: buying or adopting equipment and technology (10 programs), training existing staff (27), hiring new staff (22), expanding or relocating a facility (14), and selling abroad (11, covered in the next section). These are general-purpose economic-development instruments that a technology company qualifies for like any other employer — not technology programs. That is exactly why they go unclaimed by software companies, which assume "grants" means research money.

Equipment and technology adoption

Ten open programs pay a technology company to buy, install or evaluate technology rather than to invent it, with published ceilings from $20,000 to $2,500,000. Section 179 expensing is the universal one: it lets a company immediately deduct purchased computers, servers and off-the-shelf software instead of depreciating them, at a $2.5 million limit for property placed in service after 2024. Michigan's Industry 4.0 grant pays up to $25,000 at a 50% match for production-technology adoption; Connecticut's Additive Manufacturing Voucher up to $20,000; Arizona's Advanced Manufacturing Facilities grant up to $75,000 at 1:1. Three DOE services — Onsite Energy TAPs, Better Plants and the Better Buildings Alliance — are free technical assistance rather than cash.

Twenty-seven state programs pay to train the engineers you already have

Workforce training funds are the most-overlooked money in this entire catalog for a technology company, because they pay for something every software business does anyway. GrantCompass tracks 27 open state training programs serving technology employers, with published ceilings from $8,640 to $500,000 and a median of $62,500. Ohio's TechCred reimburses up to $2,000 per employee per technology credential and runs several rounds a year; WEDnetPA in Pennsylvania pays up to $2,000 a worker and $50,000 a company. The 27 split three ways by how the money arrives. Nine reimburse part of the cost of training the company arranges itself, listed immediately below. Thirteen fund or offset a larger training effort — a negotiated project, a hiring-and-training package, or a tax credit. Five deliver the training itself, free.

Thirteen more fund or offset a larger training effort

Thirteen of the 27 state training programs fund a negotiated project, a hiring-and-training package or a tax credit rather than reimbursing an individual credential, which suits a technology company standing up a whole team. Texas's Skills Development Fund pays up to $500,000 through a partnership with a community or technical college. Arizona's Job Training Program covers up to 75% of a custom project. New Mexico's JTIP reimburses 50–90% of a new hire's wages for up to six months of on-the-job training — a hiring subsidy delivered as training money. Kansas HPIP and South Carolina's apprenticeship credit arrive as tax credits instead.

Five states deliver the training themselves, free

Five states reimburse nothing because there is nothing to reimburse: the state designs and delivers the training itself, at no cost to the employer, using its own instructors and facilities. Georgia Quick Start is the model — the state builds a custom curriculum and delivers it free to qualifying companies creating new jobs in Georgia. LED FastStart in Louisiana, Alabama AIDT, readySC in South Carolina and North Carolina's Customized Training Program work the same way. For a small software company with no training function of its own, a curriculum built to its specification and delivered free is usually worth more than a reimbursement it would have to design and pay for up front.

Twenty-two job-creation credits pay per engineer hired

Job-creation incentives pay an employer for adding headcount, and a technology company adding engineers at technology salaries usually clears their wage thresholds easily — most are written to reward above-average wages. GrantCompass tracks 22 open programs of this shape. Fourteen of the 22 pay a share of the new hires' payroll or withholding: North Carolina's JDIG returns 25–75%, Kansas PEAK up to 95%. Two are written for startups: Illinois EDGE for Startups and START-UP NY.

Eight more pay a fixed amount per job created

Eight of the 22 job-creation programs pay a stated sum per new position rather than a share of payroll, which makes them easier to model in advance and generally less generous for high-wage engineering roles. South Carolina's Jobs Tax Credit has the widest range at $1,500 to $25,000 per new full-time job, scaled by county. Tennessee's Standard Job Tax Credit pays $4,500 a job, rising to $22,500 in its highest tier. Rhode Island pays $2,500 to $7,500 a job, Mississippi $2,000 to $5,000, Georgia $1,250 to $4,000 a year, and Florida's Qualified Target Industry refund $3,000 to $6,000. Minnesota's Job Creation Fund is the largest at up to $1,000,000, awarded on performance.

Fourteen expansion and relocation packages

Expansion packages are negotiated deals rather than application forms, and they are worth knowing about only at a specific moment: when a technology company is choosing where to put a new office, lab or plant. GrantCompass tracks 14 open programs of this kind, with published ceilings from $175,000 to $5,000,000 and several — the Texas Enterprise Fund, Michigan's Business Development Program, One North Carolina Fund — uncapped and negotiated case by case. The rule governing all of them is that the incentive must be agreed before the company commits publicly to a location. Delaware's EDGE Grant is the exception in shape: a competitive award of up to roughly $175,000, decided on merit.

Capital, export and help that is not a grant

The last group of programs a technology company can reach is not grant money at all: it is debt priced better than a bank would price it, insurance and guarantees that make a lender say yes, credits that make an investor's cheque cheaper, and expert help that costs nothing. GrantCompass tracks 11 export programs, 20 debt programs, seven investor-side credits, eight programs gated on who owns the company, and seven advisory and competition programs in the open technology slice. None of it is non-dilutive research funding and none of it should be described as a grant — but for a company that is building rather than researching, this is the part of the catalog that actually applies.

Export grants for technology products

Six open programs reimburse the cost of selling a technology product outside the United States, and they are among the least demanding programs on this page to apply for: the amounts are small, the paperwork is short, and the eligibility test is essentially that the company exports. Which programs are simplest to apply for across the whole catalog is answered on easiest small business grants. The SBA State Trade Expansion Program typically awards $2,500 to $15,000 per company for trade shows, translation, compliance testing and export marketing, and is administered by each state. Oklahoma's STEP Fund pays up to $24,000 a year. Colorado's Advanced Industries Export Grant pays up to $15,000 and is written for exactly the advanced-industry companies this page serves. New York's Global NY Fund pays up to $25,000, and Oregon's export program up to $7,500.

Export finance and insurance

Export finance solves a different problem from an export grant: it covers the working-capital gap between shipping a product and being paid for it, and the risk that a foreign buyer does not pay at all. Five federal programs serve technology exporters. The SBA Export Working Capital Program guarantees lines up to $5,000,000 against export orders; the SBA International Trade Loan lends up to $5,000,000 for facilities and equipment used in export production; the SBA Export Express Loan reaches $500,000 on an expedited approval. EXIM Bank's small-business export credit insurance covers 95% of invoice value against non-payment, which converts a risky foreign receivable into borrowable collateral.

SBA-backed and CDFI debt for technology companies

Ten debt programs — four federal, six from private lenders — lend to technology companies on terms a conventional bank would not offer a thin-file software business, with published ceilings from $15,000 to $14,000,000. The SBA 7(a) loan is the general-purpose instrument at up to $5,000,000; the SBA Express Loan trades a lower $500,000 ceiling and a smaller SBA guarantee for a much faster approval. Community development financial institutions underwrite on character and cash flow rather than collateral: Kiva lends $1,000 to $15,000 at 0% interest with no fees, Accion Opportunity Fund $5,000 to $250,000 with free coaching, and LiftFund and TruFund up to $1,000,000 in the states each covers. Founders First Capital Partners provides $25,000 to $250,000 of revenue-based financing repaid as a share of monthly revenue, with no equity taken.

State loan funds that lend to technology companies

Nine states run their own lending programs open to technology companies — plus JobsOhio, a private non-profit acting for the state of Ohio — generally at below-market rates and generally requiring job creation or in-state investment in return. Published ceilings run from $100,000 to $7,500,000, with a median of $2,000,000. West Virginia's First Small Business Growth Program is the largest at $1,000,000 to $7,500,000 per company. Oregon's Business Development Fund and Illinois's Advantage Illinois Participation Loan Program each reach $2,000,000, the latter by participating alongside a private lender rather than replacing one. The JobsOhio Growth Fund lends $500,000 to $5,000,000. Hawaii's HI-CAP uses federal State Small Business Credit Initiative money to provide loan guarantees and collateral support rather than direct lending.

Investor-side credits make a priced round easier to close

Seven programs pay the investor rather than the company, which sounds irrelevant until a technology founder is trying to close a round: a state credit worth 25% of the cheque materially improves the return an angel is underwriting, and founders in those states routinely lead with it. Kentucky's Angel Investment Tax Credit is the most generous at 25–40% of the investment. Minnesota's Angel Tax Credit gives a 25% refundable credit; Connecticut and Wisconsin also run 25% credits, Wisconsin capped at $250,000 a year. Indiana's Venture Capital Investment credit is 25–30% and is certified to the company, which then markets it to investors. New Mexico's is capped at $62,500.

Programs gated on who owns the company, not what it builds

Eight programs in the open technology slice are selected on the ownership or background of the founders rather than on the technology, and a technology company that qualifies should treat them as a separate track rather than a substitute. The SBA 8(a) Business Development Program is the most valuable: it permits sole-source federal contracts up to $5.5 million for services and $8.5 million for manufacturing, which for a software company selling to government is worth more than most grants. The DoD Mentor-Protégé Program pairs a small supplier with a prime contractor. Women's Business Centers, Veterans Business Outreach Centers and Boots to Business are free. Owner-specific hubs: women-owned, veteran-owned, 8(a) certified.

Free federal advice is the most under-used resource on this page

Two federal networks give technology companies expert help at no cost, and both are chronically under-used by software founders who assume they are aimed at retail and restaurants. SCORE provides free one-on-one mentoring from experienced executives, including retired technology operators, for as long as a founder wants it. The SBA's Small Business Development Center network offers free counselling at roughly 1,000 locations, and is the practical place to get an SBA loan package assembled or a grant application reviewed before submission. Manufacturing-adjacent technology companies can add the NIST Manufacturing Extension Partnership, which provides subsidised engineering consulting through a centre in every state.

Accelerators, incubators and pitch competitions

Five open programs award money or investment through a competitive selection rather than an application review, and they suit a technology company with a demonstrable product and a story. The Wells Fargo Innovation Incubator awards up to $250,000 to built-environment and clean-technology startups and is co-administered with a US Department of Energy national laboratory, so the award carries independent technical validation — non-dilutive, and unusually large for a private program. The Entrepreneurship World Cup awards $20,000 to $200,000 equity-free. Startup World Cup carries a $1,000,000 grand prize, taken as an investment rather than a grant. gener8tor invests $100,000 for equity through a cohort accelerator, and SXSW Pitch pays in exposure and investor access rather than a disclosed cash prize.

Funding for bootstrapped and non-venture software companies

Non-dilutive funding is not only for venture-backed startups -- and in fact SBIR rules limit how much of a company can be owned by venture, hedge, or private equity firms, which means heavily VC-owned companies can lose eligibility. Bootstrapped, profitable, and revenue-funded software companies have a clear menu.

A bootstrapped software company's realistic funding stack has four rungs, in order of value per hour spent: the federal and state R&D tax credits on development already being done; state workforce money on training already being paid for; accessible debt or revenue-based financing for growth; and, only if the roadmap contains genuine technical research, an SBIR award. Working that order downward — rather than starting with the word "grant" — is what separates a company that collects money every year from one that writes one proposal and gives up.

The R&D tax credit is the bootstrapper's anchor

For a company funding itself from revenue, the federal R&D credit is usually the highest-value, lowest-friction program, because ongoing software development generates qualifying expenses every single year. If you're not yet profitable, the payroll-tax offset (up to $500,000 per year) converts that into cash; once you are profitable, it reduces income tax. State R&D credits stack on top, and several (Hawaii, Arizona) pay cash refunds.

The practical discipline for a bootstrapped company is to have the R&D credit study done by a specialist who works on contingency or a fixed fee, and to start the documentation before the tax year ends rather than reconstructing it afterwards. Contemporaneous records — project descriptions, time allocation, and what technical uncertainty each sprint was resolving — are what make a claim defensible.

Growth capital without giving up equity

Revenue-based financing is the natural alternative to a venture round for a SaaS company with recurring revenue. Founders First Capital Partners provides $25,000 to $250,000 in revenue-based financing -- repaid as a share of monthly revenue, with no equity taken -- alongside small grants and accelerator support. For working capital and equipment, CDFI lenders are accessible even to thin-file companies: Kiva offers 0%-interest microloans up to $15,000 with no fees, and Accion Opportunity Fund lends $5,000 to $250,000 with flexible underwriting and free coaching.

The trade-off between revenue-based financing and a bank loan is straightforward: revenue-based financing costs more in absolute terms but flexes with a bad month, while a term loan is cheaper and unforgiving. A software company with genuinely recurring revenue and low churn can usually get better pricing from an SBA 7(a) loan than from revenue-based financing, provided it can wait for the underwriting.

Section 179 expensing improves cash flow immediately

Free expertise is the most underused resource

Two federal resource networks cost nothing and are open to any company: SCORE provides free one-on-one mentoring from experienced executives and entrepreneurs, and the SBA's Small Business Development Center network offers free counseling at roughly 1,000 locations -- including help with loan packaging and grant readiness. State workforce programs add training money: Ohio's TechCred reimburses up to $2,000 per employee per technology credential, helping a small IT or software firm upskill its team at the state's expense. For accelerators and incubators that support the innovation ecosystem, the SBA Growth Accelerator Fund Competition awards $75,000 to $150,000.

For high-growth deep-tech ventures, larger federal programs exist as context

Beyond the small-business programs above, the federal government funds technology at far larger scale -- but these programs are built for consortia, national labs, and venture-scale deep-tech companies, not typical small businesses. Department of Energy, Economic Development Administration, and large NSF consortium awards routinely run into the tens or hundreds of millions of dollars and require matching funds, multi-organization teams, and substantial administrative capacity. A pre-seed software company should not orient its funding plan around them.

Where these larger programs do touch smaller deep-tech firms, it is usually through prize-style competitions and lab partnerships rather than direct mega-grants. The Department of Energy's American-Made Challenges, for instance, run prize competitions in the $50,000 to $3 million+ range plus national-lab vouchers -- a far more accessible on-ramp than a consortium award. The right sequence for a high-growth deep-tech venture is to use SBIR/STTR and state matches to reach a credible prototype, then pursue the larger programs (often as a subrecipient or consortium member) once the company has the scale and partnerships those awards require.

The largest single program in this slice is a loan, not a grant

The biggest technology program GrantCompass tracks is the Department of Energy Loan Programs Office's Advanced Technology Vehicle Manufacturing loan, which typically writes $100 million to $1 billion or more. It is included here for completeness and for scale calibration, not as an action item: ATVM lends to companies re-equipping US plants to build advanced vehicles and their components, and the diligence burden assumes a corporate finance team. The same is true of the NIST CHIPS R&D Broad Agency Announcement, whose $10 million floor is a consortium threshold rather than a startup one. A useful rule for reading any technology program: if the published minimum exceeds a company's annual burn, the program is describing a different kind of company.

How this page was built, what it includes, and what it leaves out

Every program named on this page comes from the GrantCompass US catalog, which held 736 US federal, state, municipal, private and foundation programs on August 28, 2026. The technology slice is the 349 programs that list technology or software among the industries they serve. A broad industry match is a starting point, not a verdict — a general-purpose state grant tagged "technology" is not a technology program — so each of the 349 was judged against one of three tests before being linked. Test one: technology by design. Eligibility or scoring turns on research, innovation, commercialising a technology, or a named technology sector. Test two: purpose-relevant. The program is general-purpose but pays for something a technology company specifically and commonly buys — engineers' training, engineering hires, equipment, exporting a product, working capital. Those appear under a heading that says plainly they are not technology programs. Test three: excluded. Anything that passes neither. 217 programs pass tests one or two and are linked above — 62% of the technology slice, up from 35 links before this revision. 196 of the 218 currently-open programs are linked; the 21 others linked here are between intakes, upcoming, or on annual cycles, and each card says so.

What this page deliberately leaves out

Exactly 22 of the 218 open technology programs were excluded, in five groups, and naming them is more useful than silently dropping them. Six commercial utility rebates — the Con Edison, PG&E, Southern California Edison, Xcel Energy, FPL and ComEd business efficiency programs — are tagged technology but pay for lighting, HVAC and refrigeration, which a software company does not buy. Two premises programs, the Chicago Small Business Improvement Fund and the City of Dallas Small Business Assistance Program, pay for buildings rather than technology. Four public-development funds — EDA Public Works and Economic Adjustment Assistance, the EDA Revolving Loan Fund, the South Dakota REDI Fund and the Wyoming Business Ready Community grant — are awarded to municipalities, development authorities and lending intermediaries rather than to operating companies. Four general microgrants — the 1517 Fund Medici Grant, the Freed Fellowship Grant, IFundWomen and the NASE Growth Grant — are open to any small business and are covered on small business microgrants. Six always-claimable federal and multi-state records — the Section 45X advanced-manufacturing, Section 45Q carbon-capture and Section 45S paid-family-leave credits, Treasury's SSBCI state capital programs, EPA Pollution Prevention technical assistance, and our advisory record for Section 174 — are permanently available rather than periodically open, and none of them turns on being a technology company: 45X pays per clean-energy component produced, 45S on leave wages paid to any employee, and Section 174 is not a program to apply to at all but the rule governing when research costs may be deducted, explained in full above. Separately, 131 of the 349 technology programs are between intakes, upcoming or ended; several of the best are named above with their status marked on the card, and none is presented as open when it is not.

Where the numbers come from

Program counts, ceilings, medians and status flags on this page are computed from the GrantCompass US catalog as of August 28, 2026, and are stated as counts of catalog records rather than as claims about the whole US funding system — no catalog is complete. Award ceilings, program rules and dates are attributed to their agencies and dated: sbir.gov/about (read August 28, 2026) for the SBIR and STTR statutory ceilings; seedfund.nsf.gov (read August 28, 2026) for what NSF's America's Seed Fund supports; and the Internal Revenue Service's guidance on the qualified small business payroll tax credit for increasing research activities (read August 28, 2026) for the Section 41 payroll offset mechanics. Where a figure is not published by the administering agency, this page says so rather than estimating. GrantCompass is an independent funding-discovery tool and is not affiliated with any government agency; corrections are welcome via the about page.

Your situation, specifically

Three profiles cover most technology companies that reach this page, and the right first move differs sharply between them. A pre-revenue deep-tech startup with real technical risk should go straight at SBIR. A bootstrapped SaaS company should go straight at the R&D tax credit and its state twin. A software or IT services firm doing client work should skip the grant layer entirely and take the training, equipment and debt money instead. Each profile below names the specific programs and the order to work them in.

The ordering matters more than the list. Every profile below leads with the highest-value program that the company is actually eligible for today, and treats the rest as parallel tracks rather than a queue — a state SBIR match, an R&D tax credit study and a training reimbursement can all run at the same time, administered by different people, without competing for the founder's attention.

Decision tree: where do you start?

If you are not doing research, the tree continues here

The remaining branches apply to every technology company, including those that answered "no" at the top of the tree. None of them requires novel research, and all four can be worked in parallel.

Frequently asked questions

These seven questions cover what technology founders most often need settled before they can act: what the main programs are, how SBIR works for software, how the R&D tax credit reaches an unprofitable company, which states add money on top of a federal award, whether a bootstrapped business qualifies at all, what a non-research services firm can still claim, and what exists for a company with no revenue yet. Each answer names the specific programs and the amounts published by the administering agency as of August 28, 2026, and each is written to stand on its own without the rest of the page. Where an answer touches a topic another GrantCompass guide owns in depth — SBIR mechanics, the federal R&D credit, or what is open today — it links there rather than repeating it.