A startup grant is money from a government agency, foundation or public program that pays for a specific project and does not take equity or need to be paid back, as long as you spend it as agreed. That is why founders call it non-dilutive funding. The catch is that grants fund projects, not companies: they come with eligibility rules, fixed deadlines, restricted costs, slow decisions and reporting after the money lands. If your startup does real research or technical development that matches an open call, a grant can pay for months of work. If you are building a straightforward app and need money next month, a grant is usually a distraction.
The startup in the worked example below is made up, and nothing here is legal, tax or financial advice.
What is a startup grant, and how is it different from other money?
A grant buys a defined piece of work. You promise to do the project, spend the money on eligible costs and report back. You keep your shares, and in many public programs your intellectual property too.
Type of money | What you give up | Paid back? | Best for |
|---|---|---|---|
Grant | Time, paperwork and control over how that money is spent | No, if you follow the terms | A specific R&D or innovation project that fits an open call |
SAFE or equity round | A share of the company | No, but investors own part of you | Growing the company as a whole |
Loan | Interest, and often a personal or company guarantee | Yes | Predictable costs you can repay from revenue |
Cloud or AI credits | Usually nothing, but they expire | No | Infrastructure bills only |
Prize or competition | Time to enter and pitch | No | A one-off boost, often with publicity |
Credits and accelerators are their own topics. If those are what you need, read our guides on getting AI and cloud credits for your startup and applying to a startup accelerator as a first-time founder.
When is a grant worth your time, and when is it a distraction?
A serious application can take weeks of founder time, and answers can take months. Count that cost first.
A grant is usually worth it when:
Your product needs real technical or scientific work where the outcome is uncertain, and the call is written for exactly that.
You can wait. Your runway does not depend on the decision.
You meet the eligibility rules without bending the company to fit them.
It is usually a distraction when:
You need cash to cover payroll or rent. Most public programs will not fund day-to-day running costs.
Your work is standard product building. Many research programs say plainly that they do not fund routine engineering.
The call wants matched funding, meaning you pay a share of the project costs, and you do not have that money.
Which startup grants are open to founders in late 2026?
Programs change often, so treat this as a starting map. Every row was checked on the official page on 29 October 2026, and anything a page did not state is left out. Read the live call before you plan around any figure.
Region | Program | What the official page says (checked 29 Oct 2026) |
|---|---|---|
United States | SBIR and STTR (America's Seed Fund) | Non-dilutive funding for small businesses with fewer than 500 employees, with 0% equity or IP taken by the government. Phase I proof of concept is listed at 6 to 12 months and $50,000 to $275,000. Reauthorized on 13 April 2026 through fiscal year 2031. |
United States | NSF SBIR/STTR | Up to $2 million in seed funding with zero equity, for deep technology. You start with a short Project Pitch before any full proposal. |
United Kingdom | Innovate UK competitions | Smart grants are closed to new applications. Current funding runs through specific competitions on the Innovation Funding Service, which listed 28 competitions on the day we checked, mostly tied to sectors or technologies. |
European Union | EIC Accelerator | Grant component below EUR 2.5 million for innovation at TRL 6 to 8, to be completed within 24 months. Short proposals can be sent at any time. UK applicants can only apply for the grant-only scheme. |
India | Startup India Seed Fund Scheme | Paid through eligible incubators to DPIIT-recognised startups incorporated not more than 2 years before applying. The portal shows a final notice that the last date for startups to apply was 31 May 2026, so check it for any new round. |
India (biotech) | BIRAC Biotechnology Ignition Grant | Up to INR 50 lakh as grant-in-aid for up to 18 months to reach proof of concept. Calls are announced on 1 January and 1 July and typically stay open for about 45 days. |
Canada | NRC IRAP | Advice and, in some cases, funding for incorporated, for-profit businesses operating in Canada with up to 500 employees. The relationship typically starts with advice and referrals before any funding. |
Australia | R&D Tax Incentive | A tax incentive, not a grant. It offsets some of the cost of eligible R&D. Announced changes start on 1 July 2028. |
Sources: America's Seed Fund on SBIR.gov, the Department of Energy SBIR and STTR page for the reauthorization, NSF's America's Seed Fund site, UKRI's guidance on specific Innovate UK funds, the EIC Accelerator page, the Startup India Seed Fund Scheme portal, BIRAC's Biotechnology Ignition Grant page, the NRC IRAP financial support page and the R&D Tax Incentive page on business.gov.au.
Should you look at R&D tax credits instead?
R&D tax credits are the other main non-dilutive option. Instead of competing for a call before the work, you usually claim through your tax filing for qualifying research costs you already spent. The rules are strict.
In the UK, for example, HMRC's guidance on R&D tax relief says a project must seek an advance in science or technology by resolving scientific or technological uncertainty, and only companies chargeable to UK Corporation Tax can qualify. Australia's R&D Tax Incentive works on a similar idea. Every country's version differs, so ask an accountant who knows your local scheme before you count on it.
Are you eligible? A five-line check
Plenty of applications fail before anyone reads the idea. Run this check on every call before you write anything:
Entity type. Does the call need a registered company? NRC IRAP, for example, needs an incorporated, for-profit business, and Startup India schemes need DPIIT recognition. If you are still a sole founder without a company, our guide on when to incorporate your startup covers that decision.
Ownership and location. Public programs often require the company to be registered, based or owned in that country, and the work to happen there.
Size and age. Employee caps, turnover caps and "incorporated within X years" rules are common.
Stage and sector. Does the call fund early research, a prototype, or scale-up? Does it name sectors or technologies?
Matched funding. Will the grant cover all eligible costs or only a share? Where does the rest come from?
If any line is a clear no, move on.
Where do you find open grant calls?
Go to the source. The official portals list live calls for free:
National portals. SBIR.gov and agency pages in the US, the Innovation Funding Service competition list in the UK, the EU Funding and Tenders portal for EIC calls, and the Startup India and BIRAC sites in India.
Regional development agencies. State, provincial and city funds are smaller but often less crowded.
University and innovation hubs. Incubators and tech transfer offices hear about calls early.
Funder newsletters and webinars. NSF runs regular webinars. Subscribe to the two or three funders that fit.
Keep every promising call in one tracker:
Call | Funder | Fit (1 to 5) | Eligible? | Amount stated | Matched funding | Deadline and time zone | Status |
|---|---|---|---|---|---|---|---|
Name of the call | Agency or foundation | How closely it matches work you already plan | Yes, no or check | As written on the call | None or the share you pay | Date, time and zone from the call | Reading, writing, submitted, decided |
How do you read a grant call before writing a word?
Read the full call and its guidance documents twice, and pull out these parts:
Scope. What does the funder want? If you have to stretch your project to fit, it will show.
Stage. Many calls use technology readiness levels (TRL). The EIC Accelerator, for example, funds innovation at TRL 6 to 8, which is past the lab and into real-world demonstration.
Eligible costs. Which of salaries, subcontractors, equipment and travel count?
Assessment criteria. The scoring headings are your outline. Copy them into your draft before you write.
Deadline and time zone. EIC short proposals are batched for evaluation on the first Tuesday of each month at 5pm Brussels time. If you are in India or the US, convert that before you plan the week.
NSF shows how specific funders get. Its Project Pitch instructions ask for four sections with character limits, from 3,500 characters for the technology innovation down to 1,750 for the team, and say it does not fund straightforward engineering or incremental product development. Each company can only have one pitch under review at a time, and a response typically takes one to two months.
How do you write a grant application that scores well?
Reviewers score against criteria, so make their job easy: use the funder's headings, answer the question asked, and put the evidence next to the claim. Most applications need these parts:
Problem. Who has it, how you know, and what it costs them.
Innovation. What is technically new and why it is hard, not a feature list.
Why now. What changed that makes this possible or urgent.
Team. Why these people can do it, and how gaps get filled.
Plan. Work packages with milestones you can prove, and the risks.
Budget. Costs tied to milestones.
Impact. Customers, jobs and wider benefit if it works.
Here is weak versus better, using the made-up startup from the worked example below:
Weak answer | Better answer |
|---|---|
"Our AI-powered platform revolutionizes cold storage monitoring." | "We are testing whether a battery sensor can detect a failing freezer seal from temperature patterns alone, without a door sensor." |
"We have a world-class team." | "Our technical lead built firmware for low-power sensors for six years. We still need a data scientist, budgeted in work package 2." |
"The market is huge." | "Twelve small warehouses we interviewed lose stock to slow seal failures. Eight agreed to host test sensors." |
"We will build the product." | "Month 3: lab test with 90 days of logged data. Month 6: field trial in four sites. Each milestone has a go or no-go test." |
The better answers name the uncertainty, show evidence and promise something checkable.
How do you build a grant budget?
Keep the budget boring and traceable, so every line matches a task in your plan.
Start from the work, not the maximum. Price each work package, then add them up.
Use real rates. Actual salaries, quotes for equipment and subcontractors, and the funder's overhead rules.
Only eligible costs. Leave out anything the call excludes.
Show your share. If matched funding is required, say where it comes from and when.
Plan cash flow. Some grants pay in arrears after you claim, so you may have to spend first and get paid back later.
Why do grant applications get rejected?
Not eligible, or a required document missing.
Out of scope for this call, however good the idea.
Routine product development presented as innovation.
A vague plan, or a budget that does not match it.
Thin evidence of demand.
Late, because of a misread time zone or unfinished portal registration.
Keep any reviewer feedback. NSF, for example, shares a brief explanation when it does not invite a full proposal, and that is your best guide for the next attempt.
How do you spot grant scams and bad grant consultants?
Where there is free money, there are people selling access to it. The US Department of Health and Human Services has a plain guide to avoiding grant scams. It says there are no fees to apply for a government grant, that it is illegal to ask you to pay to apply for or improve your odds of a federal grant, and that lines like "We'll do all the work. You just pay a processing fee." are warning signs. In India, the DPIIT startup recognition page says the ministry charges no fee for recognition and has not appointed any agency to get it for you.
Real grant writers and consultants exist and some are good. Watch for these signs before you hire one:
They guarantee you will win.
They want a large upfront fee and will not explain what you get for it.
They want a success fee paid from the grant. Check whether the funder allows that cost before agreeing.
They will not let you speak to past clients.
They contacted you out of nowhere about a grant you never applied for.
A good helper improves your draft, but you will deliver and report on the project, so understand every line.
What happens after you win a grant?
Winning is the start of a contract. Expect a grant agreement, progress reports, claims for money spent and the possibility of an audit. Some programs pay part upfront and the rest as you claim.
Set up the boring parts on day one: a separate cost code or account for the project, timesheets if salaries are claimed, and every receipt and invoice stored where you can find it in two years. If your books are a mess, fix that first with a simple routine like the one in Aura++'s guide to setting up bookkeeping as a solo SaaS founder. Tell the funder about plan changes early, not at the final claim.
A worked example: a made-up startup applies for its first grant
Everything in this section is invented to show the process. Leena and Tobias run Frostlark, a made-up two-person startup building a low-power sensor that warns small warehouses before a freezer seal fails.
Choosing the call
They listed six calls. Two needed a larger partner, one funded only scale-up, and one closed in nine days. That left a regional innovation call and a national research program that wanted a short pitch first. Eligibility checked out line by line, including a 30% matched share they could cover from savings.
Writing and waiting
They pasted the assessment criteria in as headings. Their first draft called the product "smart monitoring", which said nothing. The second named the uncertainty and the two milestones that would prove it. They priced three work packages, attached letters from four warehouses and submitted two days early. The regional call said yes after ten weeks. The national pitch was declined as too close to product development, which told them what to change next time.
A 30-day plan to apply for your first startup grant
Days | What to do | Done when |
|---|---|---|
1 to 3 | Write one page on technical work you plan to do anyway. | A project worth doing without a grant. |
4 to 8 | Search official portals and regional agencies and fill the tracker. | Five to ten calls with deadlines and time zones. |
9 to 12 | Run the eligibility check, pick one call, register on its portal. | One call you clearly qualify for. |
13 to 20 | Draft under the criteria headings and request support letters. | Full draft, letters requested. |
21 to 25 | Build the budget from the work packages. | Every line matches a task and is eligible. |
26 to 30 | Have an outsider score the draft, fix it, submit a day early. | Submission confirmation saved. |
Startup grant FAQ
Do startup grants have to be paid back?
Usually no. A grant is not a loan, so you keep the money as long as you spend it on the agreed project and follow the terms. Misuse or ineligible spending can mean returning some of it. Some programs mix grants with loans or equity, so check what you sign.
Can a pre-revenue startup get a grant?
Yes. Many grants exist for exactly that stage, such as SBIR Phase I proof-of-concept awards and BIRAC's ignition grant. What matters is that the project fits the call and you meet the eligibility rules, not that you have revenue.
Can I apply for a grant and raise a SAFE at the same time?
Often yes, since a grant does not take equity. Check the call's ownership and control rules, since some programs restrict who can own the company, and tell your investors about the grant's obligations.
Should I pay a grant writer?
Only after you have read the call yourself and know it fits. If you hire help, pay for defined work, not a guaranteed win, and never pay a fee just to apply. You still need to understand and own every line.
When your grant-funded product is ready for its first users, you can launch it on EarlyHunt.