From 24 to 1 to 5: YC is no longer investing in Crypto, but Crypto has not disappeared

I’ve been in the crypto industry for six or seven years, and over the past two years, I’ve also delved deeply into the AI space, residing mainly in Silicon Valley. Because I’m involved in both circles, one clear feeling is: in the mainstream Silicon Valley scene, the word “Crypto” is mentioned less and less, but the things built with crypto are increasingly being used.

I want to bring back some signals from the AI side for crypto practitioners to consider.

This misalignment is most obvious at YC.

YC Winter 2026 was just announced, with 149 companies, of which 5 are related to crypto. That’s not a high number, but if you look at historical data, you’ll see a clear story behind these 5.

A set of data:

Since YC started investing in crypto projects in 2014, they’ve invested in a total of 177 companies. Breaking down the numbers by batch, the trend is quite visual:

2018-2019: 3-7 companies per batch, steadily increasing.
2020: 5-7 per batch, beginning to accelerate.
2021: jumped to 13-15 per batch.
2022: peaked — Winter batch invested in 24, Summer batch in 20, totaling 44 crypto companies that year.

Then, a cliff.

2023: still 10-13 per batch, holding for a year.
2024: a collapse begins — Winter 7, Fall 4, Summer just 1. In the entire summer, YC invested only one crypto company.

2025: Winter briefly rebounded to 10, but Spring and Summer dropped back to just 2 each.

By Winter 2026: 5 companies.

If you’re a crypto practitioner, seeing “from 1 back to 5” might seem like a sign of recovery. But if you look at what these 5 are actually doing, you’ll find they are almost a different species from the 24 companies in 2022.

What were the crypto companies YC invested in during 2022? DeFi protocols, NFT infrastructure, DAO tools, L2 scaling, blockchain games, social tokens.

And what are these 5 doing in 2026? Stablecoin deposit APIs, cross-border neobanks, trading execution engines, AI agent payment gateways, attention marketplaces.

None are building chains, protocols, or anything you’d traditionally call “Crypto tracks.”

This isn’t a rebound — it’s a blood transfusion.

Three certain projects:

Let’s quickly review three relatively straightforward ones.

Unifold — a New York team building Stripe for crypto deposits. An API + SDK that allows any app to connect to cross-chain, cross-token on-chain deposits with fewer than 10 lines of code. Co-founder Timothy Chung previously worked on Streambird (wallet-as-a-service, later acquired by MoonPay to become MoonPay Wallets), and has also been at Polymarket and Instabase. Another founder, Hau Chu, graduated from Cornell Tech. It’s a typical developer tools business — users don’t need to understand the underlying crypto.

SpotPay — a San Francisco team building a cross-border neobank based on stablecoins. CTO Thomas was previously at Google and was the 4th engineer at Brex. CEO Zsika, also from Google, has a Stanford MBA, grew up in the Caribbean and Latin America, and has firsthand experience of how painful cross-border remittances can be. The product is straightforward: one account handles overseas payments, local payments, global spending (with a physical card), and savings. It runs on stablecoins underneath, but the front end looks like a fintech app — no visible crypto.

Sequence Markets — a five-person team in New York building intelligent trading execution for digital assets. They help institutional investors route trades across exchanges for better prices and lower slippage. Fully non-custodial, they don’t touch user assets, only provide technology — a classic “selling water” model.

The commonality among these three? Crypto is a pipeline, not a selling point.

Two projects worth more discussion:

Orthogonal — AI agents that spend crypto when they pay.

I think crypto practitioners should seriously look at this.

Founder Christian Pickett previously worked on payments at Coinbase and also at Vercel. Bera Sogut worked on reCAPTCHA and Maps APIs at Google, and at Amazon Robotics; he’s a two-time ACM ICPC world finalist.

Their problem: increasingly, AI agents need to call various paid APIs to complete tasks. But these agents don’t have credit cards or bank accounts, so they can’t go through the usual registration-attach-card-pay process. Currently, developers pre-fund agents or bind their own API keys. When there are only a few agents, that’s manageable, but when thousands or tens of thousands of agents need to autonomously call hundreds of paid services, the system breaks down.

Orthogonal built a unified gateway: agents connect via MCP or SDK, accessing hundreds of paid APIs in real-time, paying per request, without managing API keys or establishing billing relationships. API providers only need to list once, and all agents can discover and call them. Settlement is done on-chain using crypto, supporting the x402 protocol — an on-chain implementation of HTTP 402 Payment Required.

Why is this related to crypto? Because machine-to-machine micro-payments in real time are exactly what traditional finance struggles with — credit card fees, bank transfer delays. These frictions are tolerable in human transactions, but for agents calling thousands of APIs daily, they’re a hard barrier. Crypto’s programmability, instant settlement, and permissionless nature make it a perfect fit.

Timeline note: YC’s Fall 2025 RFS (Request for Startups) emphasized “Infrastructure for Multi-Agent Systems,” and six months later, they invested in Orthogonal. Early supporters include YC alumni like Precip (W24), Riveter (F24), Andi (W22), Fiber AI (S23), all working on agent products, indicating this isn’t just theoretical — it’s a real need.

An interesting intersection: in a recent viral article, it was said “Agents are the new masters of software,” and SaaS is shifting from B2B/B2C to B2A (to agents). If that’s true, then payments between agents become a fundamental infrastructure problem — and Orthogonal is betting on crypto to solve it.

Forum — turning “attention” into a tradable asset.

This project is the most imaginative, but also the riskiest.

Founder Owen Botkin previously traded long/short equities at Balyasny, one of the top hedge funds globally. Joseph Thomas has worked at NASA and DreamwaveAI. YC’s partner for this project is Jared Friedman — one of YC’s core partners.

Forum aims to be “the first regulated attention marketplace.” Specifically: building indices from data in search engines, social media, streaming platforms to quantify how much attention a topic, brand, or cultural phenomenon is getting, then allowing users to go long or short on changes in that attention.

For example: if you think a brand is about to lose public attention due to a PR crisis, you can short its attention index. If you believe a cultural trend is heating up, you can go long.

Their core argument: attention is the primary driver of business success in the digital age — advertising, traffic, user growth — all are monetizations of attention. But attention itself has never been directly priced or traded.

This project isn’t labeled as crypto/Web3 now, but “regulated exchange” plus “creating a new asset class” strongly suggests tokenization. In YC’s Spring 2026 RFS, the phrase “new financial primitives” first appeared, alongside AI-native workflows and modern industrial systems.

This is YC’s first explicit crypto-related topic in recent years. The wording is specific — not just “blockchain” or “Web3,” but “stablecoin financial services,” with concrete directions like yield-bearing accounts, tokenized real-world assets, cross-border payment infrastructure.

My take:

As someone active in both crypto and AI, I see this data as good news — but perhaps not in the way many expect.

YC hasn’t abandoned crypto; they’ve redefined what kind of crypto companies are worth investing in.

In one sentence: YC is no longer investing in crypto, but investing in companies that use crypto.

The difference? The former’s value proposition is “building the crypto ecosystem,” while the latter’s is “solving real problems with crypto as the most suitable tool.”

The former’s users need to understand wallets, gas fees, on-chain interactions. The latter’s users often don’t even realize they’re using crypto — SpotPay users think they’re using a bank app, Unifold clients think they’re integrating a payment SDK, and Orthogonal’s agents don’t even have the concept of “thinking” about it.

What does this mean for us?

First, good news: stablecoin payments have shifted from niche consensus to mainstream Silicon Valley recognition. YC’s dedicated RFS topic, the progress of the GENIUS and CLARITY acts, Stripe’s acquisition of Bridge — all signals that the regulatory path for stablecoins is opening. For teams deeply involved in this space, funding and market perception are improving.

Second, new opportunities: agent payments are a demand emerging from within the AI industry. Crypto practitioners have a natural advantage here. Real-time machine-to-machine micro-payments, programmable money, permissionless settlement — these have been discussed for years, and now they have a concrete application in the agent economy. It’s not us seeking a scenario — the scenario is seeking us.

Of course, there are realities to face: the profiles of competitors are changing. CTO of SpotPay was the 4th engineer at Brex; founders of Orthogonal come from Coinbase and Google — not crypto-native backgrounds, but with traditional tech company engineering and product methodologies. Competing with them requires more than just understanding chains; we need to improve product experience and engineering.

Also, directions like L1/L2, DeFi protocols, NFTs, DAO tools — not valueless, but in the eyes of Silicon Valley accelerators and VCs, no longer top priority. That doesn’t mean these are dead, but if you’re working in those areas, your fundraising strategy and narrative may need adjustment.

Finally, the “24→1→5” data trend: I believe the most accurate interpretation isn’t “crypto is recovering” or “crypto is declining,” but that crypto is being redefined.

YC spent two years figuring out one thing: the biggest value of crypto might not be as an independent industry, but as infrastructure for other industries. Whether this judgment is correct remains to be seen, but as someone active in both tracks, I see a lot of opportunities for crypto practitioners — provided we’re willing to see ourselves from a different angle.

Crypto doesn’t need to disappear, but the best crypto products might be ones users don’t even realize involve crypto.

That’s not compromise — it could be the greatest victory.

You may disagree with this view, but this is the stance of the most influential startup accelerator in Silicon Valley, expressed with real investment.

Data sources: YC Directory (Crypto/Web3 tags, all 177 batches), YC Winter 2026 Launch List (149 companies), YC Request for Startups (Summer 2025 / Fall 2025 / Spring 2026). Details of five crypto-related projects are from YC’s official site and public info of the companies.

Author: aiwatch, six years in crypto, recent two years deep in AI, based in Silicon Valley, focusing on GenAI product analysis and cross-field research of Crypto×AI.