Chinese and American AI companies rush to go public; crypto can only wait for opportunities in the corner.

By: Connor Dempsey

Translation: Deep Tide TechFlow

Deep Tide Editor’s Note: Connor Dempsey is a veteran in the crypto industry. He has worked at Circle, Messari, and Coinbase Ventures, and currently handles marketing at Crossmint. In this short commentary, he puts forward a view: under China’s “common prosperity” logic, China is pushing AI companies to go public quickly at reasonable valuations, while U.S. peers have to wait until the end of 2026 to IPO, at which point valuations could be 100 times those of China’s companies. The AI IPO boom will continue to siphon away market capital and put near-term pressure on crypto, but 2026 may be a good year for early investors.

Body:

A wave of IPOs from Chinese AI companies is coming. It will last 1–2 years.

And their valuations are more attractive than those of U.S. peers.

Preliminary note: The following is from a conversation I had with a longtime friend who is familiar with the Chinese market (for me, the Chinese market is basically a black box).

The logic of China

China’s AI buildout speed is as fast as the U.S.’s. But China has stronger control over the private sector; what they worry about is the gap between the rich and the poor.

The logic is like this: AI is a winner-takes-all game. The longer companies stay private, the more wealth concentrates in the hands of a small number of founders and investors.

Chinese policy puts pressure on high-growth AI companies, pushing them to list earlier so ordinary investors can also share in the gains.

MiniMax (AI video generation) and Zhipu (China’s OpenAI) have already gone public. Mianzhisuo (chatbot), Baichuan Intelligence (medical AI), and Kunlunxin, which is under Baidu, are in line. Their IPO valuations are between $2 billion and $7 billion, which seems reasonable. DeepSeek is the only exception, claiming it will keep continuing with private financing.

As a Westerner, I’m not trying to take China’s side, but this line of thinking does make sense. U.S. AI giants won’t distribute the same level of wealth to the public.

The U.S. timeline

The U.S. also has a wave of AI IPOs, expected to start in late 2026 to early 2027. OpenAI, Anthropic, Databricks, Perplexity, and Elon Musk’s xAI (and after it merges with SpaceX) should all be listed within this time window.

But by the time ordinary investors are able to buy, those companies’ valuations could be 100 times those of their Chinese counterparts.

When OpenAI and Anthropic go public, they will likely already be valued at the trillion-dollar range. Databricks and xAI are already above $100 billion today.

AI is the only card on the table

No matter how it ultimately evolves, the China–U.S. AI race will be pushed forward at full speed, and it may suck the air out of every other technology track, including crypto.

Why? Because AI is the most important technology of our lifetimes. If you’re a capital allocator, it’s hard to look in other directions right now. For example, if you have $1 million to deploy, you’ll most likely find a way to ride this AI wave.

As long as the AI IPO feast keeps going, crypto asset prices will likely face downward pressure.

Venture funding in crypto has already visibly slowed down. I recently met a founder who raised funding in the crypto space, and his observation was: if you’re not doing AI, most investors have zero interest.

The bright side

It’s not new for crypto to see sentiment hit rock bottom and for investors to lose interest. For example, after the ICO bubble burst in 2018, most investing the public turned cold on crypto for a full two years.

But if you were an early investor during that period, you did extremely well. Solana, Compound, and Uniswap’s seed rounds were funded back then. Circle’s USDC (2018) was also built in that same timeframe.

I think for those who are still deploying early capital in crypto, 2026 could produce a similarly great year.

Meanwhile, U.S. crypto regulation is becoming clearer, and infrastructure work around tokenizing and overhauling financial market systems is underway.

Protocols like Hyperliquid have already started spilling over into traditional markets, offering 24/7 crude oil futures trading when traditional markets are closed on weekends (see Syncracy’s “The Great Perpification”).

Although there are fewer crypto founders now, there are still new and interesting companies launching. For example, Ryan Yi—who brings four years of venture capital and corporate business development experience from Coinbase—founded Onchain Group. In essence, it’s an investment bank built for the token economy. You can think of it as traditional M&A, but with the deal objects swapped for tokens, and the customers being the largest protocols in crypto.

AI is the variable for crypto

Even though AI is siphoning away the room’s air at the investment level, ultimately it will become rocket fuel for crypto’s practical use.

Crypto has always had user experience problems. By combining an AI front end with a crypto back end, using crypto can suddenly become as easy as using Claude or ChatGPT. In the end, this will become a massive entry point for crypto assets and protocols.

AI Agents are also expected to become crypto’s largest user group. The business between Agents—millions of AI Agents autonomously trading without human involvement—is one of the strongest use cases for stablecoins and blockchains. When you need millions of Agents to create wallets and trade at a fraction of a cent, traditional card networks will collapse. Crypto won’t.

I’m betting that the fusion of crypto and AI will be a big deal in its maturity phase. It’s also one of the most worth-watching phenomena in this industry’s history.

~CD