On Thursday night, October 1, Wu Blockchain's monthly venture report landed in my inbox while I was halfway through a pitch deck. The headline was cheerful: crypto VC funding rose 71.1% in September to $1.269 billion across 61 deals, according to RootData figures.
I read the top-ten table twice. Then I went back to the deck, which opened with a slide about a token, and I felt a small pang for the founder who'd made it.
Here's my read, plainly. The stablecoin fintech funding that carried September went to companies whose customers would be surprised to hear they use crypto at all. The cheques followed the bridge, and the bridge was built to be invisible.
What the September Numbers Actually Say
A 71% month-on-month jump sounds like a recovery until you check the base. August was $742 million, and September 2025 was $6.096 billion, so last month is still down 79.2% year on year. Deal count barely moved. There were 62 deals in August and 61 in September, which TokenPost's breakdown puts at 35.1% fewer than a year earlier.
So the rebound is a handful of large rounds rather than a broad thaw, and if you're a seed-stage founder wondering why your inbox didn't feel 71% warmer last month, that's the reason — the median cheque didn't move, only the tail did. Polymarket alone took $300 million from 1789 Capital at a $21 billion valuation. That's almost a quarter of the month.
Take Polymarket out and look at what's left. Felix Pago raised $200 million, Jeeves raised $110 million, and HIFI raised $37 million. Together that's $347 million, or about 36% of the non-Polymarket total. None of the three sells crypto to its users.
I'll be honest about the size of the sample. One month is noise, and a single $300 million cheque can swing any chart. But I've watched the same shape forming in my own meetings since early summer, and September is the first month where the public numbers say it out loud.
Who Got Funded, and What Their Customers See
Felix Pago is a remittance company that runs inside WhatsApp. A worker in Houston chats with a bot, sends dollars, and her mother in Guadalajara picks up pesos. Under the hood, Felix settles in USDC on Stellar and uses Bitso as the on- and off-ramp. The customer never touches a wallet. Andreessen Horowitz led the $87 million equity portion of the round, and the company says it has processed more than $8 billion for over six million users across 11 countries.
Jeeves sells corporate cards and spend management to companies like BMW and Lululemon. Its Series C-1, led by CoinFund, brought in Coinbase Ventures alongside AllianceBernstein and Singapore's GIC. Its stablecoin business is already running at $1.5 billion a year in volume, yet a finance team in São Paulo using Jeeves sees nothing more exotic than a card, a dashboard and a monthly statement — the fact that settlement happens in USDC and USDT is a line in the vendor documentation that most of those teams will never read.
HIFI is the most explicitly infrastructural of the three. Its pitch, as reported on September 24, is one API that moves value across stablecoins, bank rails and card networks. A developer integrating HIFI is buying settlement, not ideology.
Then look at who wrote the strategic cheques, because the names on that side of the table tell you more about where this is heading than any of the venture firms do. Nasdaq put $100 million into Kraken's parent, Payward, to push tokenized equities. Four days later, S&P Global led an extension of Kaiko's Series B to $110 million. Market operators and data incumbents were buying the plumbing.
The Bridge Thesis, Applied to a Funding Table
I've said for years that the next wave of Web3 adoption won't come from better technology alone, but from better bridges between ecosystems, communities and real-world value.
What September adds is a quieter corollary. The best bridges are the ones people cross without noticing there was water underneath.
A remittance customer doesn't care whether her transfer cleared on Stellar or SWIFT. She cares that it arrived before dinner and cost less than last time. The founders who internalised that early stopped pitching the rail and started pitching the arrival. And so the investors who can underwrite a payments business, the ones who know CAC curves and interchange and fraud loss, finally had something they could price.
That's the bridge in the funding table. General Catalyst's credit fund sits behind Felix's $113 million debt line. AllianceBernstein and GIC sit next to CoinFund in Jeeves. Those names don't show up for a token thesis. They show up for unit economics that happen to run on a cheaper settlement layer.
There's a longer history here that anyone who worked in payments before crypto will recognise. Nobody who taps a card in Tel Aviv thinks about the acquiring bank, the card scheme, the issuer processor or the overnight netting file that settles the whole thing, and for most of the 2010s the fintechs that won (the ones that became household names rather than conference sponsors) were the ones that wrapped that machinery in an interface simple enough to forget. Stablecoins are going through the same wrapping now, only faster, because the rail underneath is programmable and the rules for dollar stablecoins in the U.S. have become clear enough for a CFO to sign off on them. When I look at the September table, I don't see crypto losing its identity. I see it graduating into the layer of finance people stop talking about precisely because it works.
I wrote in Bridge Notes #11 that the business in agent payments would be whoever bridges the card rail and the stablecoin rail. September looks like the human version of the same trade.
Where the Analogy Breaks
This isn't the whole market, and I don't want to pretend it is.
Polymarket is the obvious counterexample. It's the biggest round of the month, and it's unapologetically onchain, with the market itself as the product. Prediction markets don't hide the chain; their users come partly because of it. If your company's value depends on open, composable state, invisibility isn't a strategy you get to choose.
The headline number also flatters the trend I'm describing. Felix's $113 million is credit, USD.AI's $40 million is a debt facility, and World Assets' $49 million came from OTC token sales. Strip those out and September's equity venture total is closer to $1.07 billion. Credit lines go to businesses with predictable receivables, which is exactly the fintech profile, so the "invisible crypto" share is partly an artefact of how lenders behave.
And there's a real risk in invisibility. When the chain disappears from the interface, it can disappear from the founder's mind too. Some of these companies will eventually swap USDC on Stellar for a bank partner's tokenized deposit, and the customer won't notice. That's fine for them. For anyone who believes open rails matter, it's a reminder that "invisible" and "replaceable" sit closer together than the pitch decks admit.
What Should Founders Raising Now Do With This?
If you're raising in Q4, the practical lesson is about the order of your story, not about hiding what you've built.
Lead with the customer outcome and the number that proves it: transfer time, cost per transaction, payout coverage, revenue per account. Put the chain on the slide that explains your margin. In the decks I've seen land this year, crypto earns its place as the reason the economics work, not as the reason to care.
Kima Network, in our portfolio at PRIM3, is a useful example of the framing. The pitch that resonates with banks and payment firms isn't "cross-chain." It's settlement between blockchain and traditional rails, with the chain as an implementation detail they can audit.
It also pays to build for the investors who just showed up. Credit funds, strategic corporates and growth investors ask for cohort retention, loss rates and regulatory posture well before they ask about tokenomics. If you don't have those answers, September's capital isn't yours yet.
Keep the parts of your product that are deliberately visible, though. If openness, self-custody or composability is the moat, say so and price it. The worst position is a half-hidden chain that neither crypto-native users nor fintech customers can see the value of.
I made a version of this argument in Bridge Notes #14, about headline capital and the small share that reaches founders. September sharpens it. The money is back for builders whose customers don't know they're using crypto, and that's not a betrayal of the industry.
It's what arrival looks like.
FAQ
How much crypto venture funding was raised in September 2026? According to RootData data reported by Wu Blockchain on October 1, 2026, crypto VC funding reached $1.269 billion across 61 deals in September, up 71.1% from August but down 79.2% from September 2025.
Why are VCs funding stablecoin fintech companies? Because they look like payments businesses with measurable unit economics. Felix Pago, Jeeves and HIFI use stablecoins to settle faster and cheaper while their customers see a familiar remittance, card or API product.
Should a crypto startup hide the blockchain from its users? That depends on where the value sits: when the chain lowers cost or speeds settlement, put the outcome first. If openness or composability is the product, as with prediction markets, the chain should stay visible.