I started this research with a simple question: where can a founder go before the normal startup fundraising process makes sense?
Not after the deck, company, co-founder, MVP, and first customers. Before some or all of them exist.
I compared 30 accelerators, programs, residencies, and day-zero funds.
Then I tried to compare their economics.
The offers could not be compared by category or advertised amount alone.
At this stage, an offer can mean an immediate investment, a check conditional on an investment committee, an equity-free grant, a residency, cloud credits, introductions, workspace, or no direct capital at all.
The headline amount cannot tell you whether the offer is expensive. First you have to ask what is actually guaranteed, what dilution is known, and what the program creates that the founder does not yet have.
12 ultra-early-stage options and their terms
- Antler Singapore
- Antler MENAP
- Entrepreneurs First
- South Park Commons Founder Fellowship
- Village Global
- EDB Global Founder Programme
- Forge Residency
- Hub71 Initiate
- Canonical
- Antler Disrupt
- a16z Speedrun
- HF0 and Neo Residency
HF0 and Neo are separate programs. They shared one comparison row in my internal table, so the public list contains 12 rows but 13 program names.
The shortlist also contains offers with very different structures.
Entrepreneurs First says founders may apply with an idea, a co-founder, or neither. Its public site describes an equity-free grant during ideation and investment of up to $250,000 after a company is formed, but it does not publish the complete investment terms on the main page.
South Park Commons says its Founder Fellowship can start before an idea or co-founder. Its current published offer is $400,000 for 7% on a standard SAFE, plus $600,000 committed to the next external round.
Canonical is not a cohort. It describes itself as pre-seed, pre-product, and pre-deck, with $500,000 to $1.5 million first checks for technical founders building post-AGI infrastructure.
Forge is explicitly a -1 to 0 residency, but its public site does not expose standard investment terms. Hub71 Initiate offers company-building support, a license, and workspace, but says Hub71 itself provides no direct funding in that program.
Upfront investment, conditional investment, grants, and non-cash support must be recorded separately.
Compare the guaranteed money, not the marketing total
Consider Antler Singapore. Its current published structure is $100,000 for 10%, plus $50,000 on an uncapped MFN SAFE. The fixed component implies a $1 million post-money value, but you cannot divide the full $150,000 by 10% and call it a $1.5 million valuation. The second instrument creates additional dilution whose final percentage is not yet known.
Antler Disrupt uses a similar split at a different stage: $250,000 for 10%, plus $150,000 on an uncapped MFN SAFE, and only for companies that pass the investment committee after the sprint. Its headline is $400,000, but the capital is conditional and the fixed 10% is not the entire eventual cost.
a16z Speedrun publishes "up to $1 million," but the standard deal is $500,000 for 10% upfront and another $500,000 in the next round within 18 months. Again, guaranteed now and committed later are not the same thing.
This is also why the common description of the YC deal as "$500,000 for 7%" is wrong. YC invests $125,000 for a fixed 7%, then $375,000 on an uncapped MFN SAFE. The effective dilution is greater than 7% and depends on later financing terms.
What does an ordinary pre-seed round cost?
Carta's 2025 US data gives a useful size-matched benchmark. The median post-money SAFE cap was about $10 million for rounds between $250,000 and $1 million, and about $15 million for rounds between $1 million and $2.5 million.
This does not create one universal "fair" dilution number. A $200,000 angel SAFE and a $2 million institutional pre-seed are different transactions.
It also does not make an ultra-early-stage offer automatically expensive. The relevant counterfactual is not "What could an already fundable company raise?" It is "What can this founder credibly create or raise today?"
Suppose a program takes 7% before the idea exists. That may look expensive next to a later SAFE with a high cap. But if the program helps form the team, choose the market, incorporate, build the first product, and reach investors, the later SAFE may not have existed without it.
The extra dilution may pay for co-founder formation, market selection, incorporation, product development, and investor access, not only cash.
The opposite is also true. A familiar brand and a large headline package can be poor value when most capital is conditional, the founder already has the missing network, or the program requires a relocation and full-time cohort that delays the actual business.
Compare what you receive now with what you give up
Treat only cash available now as cash. Keep investment conditional on an investment committee, matching capital, or a later round separate. Count fixed equity and uncapped SAFEs separately, then add fees, relocation, incorporation, and time.
After that, compare the offer with the round the company could realistically raise today - not after an imagined future MVP.
Early capital makes sense when it creates something missing: a co-founder, a product, customer access, a location, or a credible financing path.
If you already have those things, building the MVP first may let you raise more money for a smaller percentage.
Sources checked July 26, 2026: Carta State of Pre-Seed 2025, Carta State of Pre-Seed Q1 2026, YC Deal, Entrepreneurs First, South Park Commons Founder Fellowship, Canonical, Antler Singapore, Antler Disrupt, and a16z Speedrun. Program terms change; verify them before applying.