“Can I raise with just an idea?” is the question we hear most, and the honest answer is: some money, from a few specific places, for a specific purpose. Not a seed round, not a valuation, not a year of runway. What is realistic at idea stage is enough capital to turn the idea into something other people can evaluate. That is worth more than it sounds, because it is the step most founders try to skip.
This is the guide we would want to read before asking. It covers what you can actually raise, what to prepare, the mistakes that turn an early “maybe” into a “no”, and how our own grant-to-round path works.
What is realistic before a product
- Grants sized to a first milestone. For most teams this is the most accessible idea-stage money in web3, because a grant program pays for a deliverable instead of buying a stake. On our Web3 Grants Tracker, three programs target early work directly: Superteam Earn Grants on Solana (up to $10K per grant, micro-grants up to $500, idea or prototype stage), Starknet Seed Grants (up to $25K in STRK for an MVP or proof of concept) and our own program ($5K–$75K, idea or prototype stage). Others offer small tickets that suit a first milestone inside their ecosystem: Filecoin Foundation’s Next Step Grants ($5K–$10K), the smaller tiers of Lido’s LEGO program (up to $1K or $10K) and Stellar Community Fund Instawards for small pieces of work. Programs open to any stage, such as the Ethereum Foundation’s ESP or Aleo’s developer grants (up to $100K, milestone-based), will read an idea-stage application too, as long as the milestone is concrete.
- A small first check from a fund that says it does idea stage. Most funds wait for a product. The few that do not usually say so plainly, and their first check tends to be small and tied to evidence rather than to a valuation built on a deck. Ask directly: “Do you invest before there is a product, and in what form?” A clear yes or no saves both sides weeks.
- Hackathon prizes. A prize rewards a working demo built against a deadline, so it is less a funding source than a forcing function: you leave with a demo link and, often, feedback from people who know the ecosystem. Treat hackathons as the cheapest way to produce the prototype a grant committee wants to see, not as a funding plan.
What is not realistic yet
Some money is structurally out of reach before you ship, and knowing which saves applications. Retroactive programs such as Optimism Retro Funding pay for impact already delivered. Traction-gated programs such as Prezenti on Celo (live with usage) or Aptos Foundation Grants (live product with early traction) ask for users you do not have yet. Quadratic-funding rounds weigh how many people back you, which is hard before anyone has used the thing. And a priced round on a deck alone is not something to plan around. None of these are closed to you forever; they are the second and third stops, not the first.
What to prepare: four things, one page
- One milestone. A single, checkable deliverable with a date: an MVP on testnet, a working integration, a benchmark someone else can reproduce. If a reviewer who has never met you could not verify it on that date, it is not concrete yet.
- A budget that reads like a bill of materials. Three to six line items (engineering time, review or audit, infrastructure) that add up to the ask. The total should sit inside the program’s stated range; if it does not, resize the milestone instead of bending the numbers.
- A demo or an architecture doc. Before a product, this is your evidence. A clickable prototype is best. Failing that, write a document that shows the components, what runs on-chain and why, and where the hard part is. A repo with a README that actually runs beats any deck.
- The team, with links. Who builds it and what each person has shipped before: repos, past projects, deployed contracts. At idea stage the team is most of what a reviewer is evaluating, so make it easy to check.
Our one-page grant application template on GitHub puts these in the order committees read them: a specific thesis, a reason it is on-chain, evidence you can ship, and a milestone the money completes. It is CC0-licensed, so fork it and use it for any program, not only ours.
Common mistakes at idea stage
- Asking for runway. “Twelve months to build” is a calendar, not a milestone. Committees fund something they can check off.
- Over-asking. Requesting the top of a program’s range for a first step tells the reviewer the work has not been scoped. Size the ask to the milestone, not to the program’s ceiling.
- Token first. Leading with token design before anything works reads as the reason the project exists. Put the product first; tokenomics belongs later.
- No reason to be on-chain. If the product would work the same on a normal server, say what breaks without the chain, or expect a no.
- One application, ten programs. Every program publishes what it funds, and generic text that ignores it is easy to decline. Keep the core one-pager and rewrite the fit paragraph for each program.
- Waiting to look fundable. Polishing in private for months burns the time a small grant would have bought. An honest early application with one milestone is stronger than a late one with ten.
How our grant-to-round path works
ZK Prime Capital is a private, full-cycle AI venture fund based in Singapore, focused on on-chain AI agents, DeFAI and zero-knowledge infrastructure. For us idea stage is a valid stage, and the path has four steps.
- Grant. $5K–$75K, non-dilutive, for idea- and prototype-stage teams, tied to one first milestone. Apply with one link and one paragraph, no deck. Every application gets a reply within 5 business days.
- Round. If the milestone ships and both sides want to continue, the fund can invest $50K–$250K in a priced round. It is a separate, later decision, and the first point where equity or token terms appear at all.
- Incubation. Tokenomics, go-to-market and launch preparation, worked through alongside the team.
- TGE. A token launch through the fund’s launchpad. At the fast end, idea to TGE takes about 90 days, though not every team should move that fast.
So far $3.1M in grants has gone to 60+ teams this way. The fund behind the program manages $48M, has 37 portfolio companies and has run 14 token launches. What we fund at idea stage, and what we usually pass on, is on our idea-stage funding page.
Have an idea and no product yet? Write down the milestone, the budget and the architecture, then send us the link. Apply with your project ▸