A web3 grant is not a gift. It is a chain, a protocol or an infrastructure vendor paying you to do something they need done, on terms they wrote. Business Chain tracks 27 programmes open to web3 teams right now, and the ones that match a given founder's stage and stack usually number six or seven, not 27. Pick the wrong one and you spend three months building for an ecosystem that never had the users you assumed were there.
What the funder is actually buying
Fireblocks Blockchain Grants for Web3 Innovation & Ecosystem Growth caps its grant at 25% of the client's annual fees already paid to Fireblocks. That is not generosity, that is a loyalty rebate dressed as a grant. Fireblocks is buying retention from teams already on its infrastructure, not acquiring new ones.
Base Builder Grants pay 1 to 5 ETH per project. BNB Chain Grants go up to USD 200K per project. Both are buying the same thing: transaction volume and developer mindshare on their chain, measured in dapps shipped and users onboarded, not in a polished pitch deck. If your product can live on any chain, go where the cheque is biggest relative to what they will demand back later.
What one-chain money costs you later
Starknet Seed Grants pay up to $25,000 in STRK, not stablecoin. Take that and your treasury is now exposed to STRK price movement until you convert it, and converting a grant token publicly can read as a vote of no confidence in the chain that funded you. The same logic applies to Trailblazer 2.0, a $1M pool for agentic DeFi on Arbitrum capped at $10k per submission. Ten thousand dollars will not pay for a rewrite if you later decide Arbitrum was the wrong base layer, and by then your contracts, your audits and your community are already anchored there.
The real cost of single-chain money is not the token volatility, it is the migration you will not do once you've taken it. A $10k grant with a chain's name on your audited contracts is a much smaller check than the six-figure rebuild to move off that chain a year later.
How milestone payment actually runs here
Almost none of these programmes wire the full sum on day one. Flare Grants Program releases funds per milestone under a grant agreement, and the total is not published upfront because it is negotiated against what you commit to ship. Get a Grant to Build on Conflux disburses in stages tied to milestones the same way, with no public cap either. Treat the unpublished number as a negotiation, not an omission. You propose the milestones and the dollar figure per milestone; the foundation counters.
Circle Developer Grants work the same way from the funding side: the amount is set by tiered USDC funding tied to meaningful usage, applications run on a rolling basis, and the first cohort of 2026 has already been selected. If you are reading this now, you are applying for the next cohort, not the current one, so build your milestone proposal around that timeline before you submit.
Builder programme or marketing budget. Tell them apart before you apply
A builder programme asks what you are shipping and pays against delivery. A marketing programme asks how big your audience is and pays for the association. SSV Network Grants Program, at $5M, and Alchemy + Arbitrum Grant, up to $500K in credits per team, both sit on the builder side: infrastructure credits and protocol funding scaled to teams actually running code against their stack.
The Graph grants program is worth knowing even though there are no open RFPs at this time, because its model, grants awarded from the community treasury against specific requests for proposals, is the clearest version of a builder programme in this list: no RFP, no money, full stop. dYdX Ecosystem Development Program assesses applications against stated criteria including relevance to the protocol, which is another way of saying they will fund you if you extend dYdX itself, not if you use dYdX as a case study in your own marketing.
- If the amount is capped as a percentage of fees you already pay a vendor (Fireblocks, 25%), it is a retention tool, apply once you are already a paying customer, not before.
- If the figure is a flat per-project range (Base, 1 to 5 ETH; BNB Chain, up to USD 200K), it is competing for new builders, apply early in your build, not after launch.
- If the total is unpublished and tied to milestones (Flare, Conflux), write your own milestone schedule before you apply, because you are setting the pace of the negotiation.
- If there is a hard per-submission cap (Trailblazer 2.0, $10k), use it for a specific feature or integration, not as your main runway.
- If the programme states assessment criteria tied to protocol relevance (dYdX) or requires an open RFP (The Graph), read the current RFP or criteria list before you write a word of your application.
Pick the chain for the roadmap you already have, not for the grant on offer. The grant should pay for work you were going to do anyway; if it is changing what you build, you are building for the funder, not for your users.