Meet the Most Active CVCs of 2025

"Now we know who to hit up for gold"

Bottom line first: the most active player is Ant Group.

This claim comes from my realization that Ant has spent the past year scooping up stakes in an extraordinary number of star AI applications.

The most notorious is Vivix AI's Liu Yu, at a post-money valuation of $1.32 billion — a miracle by any standard in these three years of AI frenzy. We mere mortals can't judge; we'll wait for an actual product launch and assess it on the merits. (Updated: this round may have been Alibaba's strategic investment arm instead.)

It co-invested with Sequoia in Lovart when the company was valued at $500–600 million — a true act of devotion to the Chen Mian cinematic universe.

It backed Natural Selection and another AI gaming standout valued at 1 billion RMB. I'm all for this; investing in games is simply good taste.

It also threw in for Rockflow, which makes strategic sense given potential business synergies. Though Rockflow currently only accepts existing US/Hong Kong stock traders, so here's hoping it becomes the Futu NiuNiu of its category.

Then there's the recently announced "largest Coding Agent financing round in China" — Atmos, where Ant put in 100 million RMB, plus Looki, Deep Principle, and a string of robotics projects.

Most worth mentioning is Macaron — congratulations to this old friend of the "burying AI" beat on yet another funding round, one step closer to Ant acquisition. Can I dare to hope Kaijie will save Lingguang? 😭

Ant Investment: you show up, they deliver. Never disappoints.

But Ant wasn't always like this.

Ant's two most famous investment cases: first, Orbbec, a 3D sensing and AI vision company that partnered with Alipay to launch facial-recognition payments.

Second, its bet on ofo. When ofo and Mobike looked likely to merge, Ant quickly pivoted to support Hello Inc. — providing credit-payment infrastructure and risk-control capabilities that enabled universal deposit-free rides almost overnight.

This is the core function of Ant's CVC: "using investment as a method to do business for the company."

The latest example of Ant-style investing: its 2024 acquisition of Haodf.com for 2 billion RMB. Keep in mind Haodf's shareholders included Baidu and Tencent, and rumors swirled that one major factory refused to sign off — yet Ant still closed the deal. This is the kind of fight Ant's strategic investment team should be picking 👍

Haodf.com brought Ant's "Afu" 280,000 professional doctors and associated consultation data. Finance, insurance, healthcare — all in Ant's wheelhouse. Health is itself a critical piece of the insurance business, so the all-in-on-Afu logic tracks cleanly.

I support buying ads on WeChat from Allen Zhang Only someone absolutely desperate with anxiety would cook up something like Lingguang — a product with zero connection to their core business — and still boast about building the "Alipay" of the AGI era. I'm genuinely curious how Macaron and Lingguang strategically synergize with Ant 🤓

I suspect it's pure desperation to prove they still matter. I still remember: in 2019, Ant declared it would no longer invest in pre-Series C companies. Then AI arrived, and everyone forgot who I was 😭

I'll roughly divide the past year's CVCs into two categories: those with a base, and those without.

Alibaba and Tencent both have bases. They can invest with abandon.

Alibaba has cloud services. Its investments in MiniMax and Moonshot AI, for instance, partly came as "merit coupons" paid in compute credits — and the model companies then turn around and buy cloud services. MiniMax purchased $58.3 million in Alibaba Cloud services in the first nine months of 2025 alone. This investment model is literally can't-lose.

Tencent is a traffic gateway. WeChat plus gaming channel advantages let it play purely financial investor. It can absorb AI applications valued above $500 million — Manus and Genspark, for example. I understand Tencent is also looking at Lovart and OiiOii. Wouldn't be surprised if we someday generate AI manhua dramas right in WeChat mini-programs.

Those without a base are obviously more anxious, investing while scrambling to build new businesses — messy, lacking coherent strategy.

Take Xiaohongshu: rather than hosting indie developer competitions to harvest traffic, maybe try nurturing more humanities majors to become AI influencers? Though honestly, I don't go on Xiaohongshu for AI content anyway.

As for Xiaohongshu investments, my first impression is they've picked up some ZhenFund hand-me-downs — I understand they're looking at Typeless. My second impression is their splash in "stirring up Shenzhen's investment circle" last year, backing a "flying pet that defies physics." Though insiders insist this thing actually flies. Anyone seen it?

Then there's JD.com. Fu Zai and Looki both launched on JD. Fu Zai supposedly sold 120,000 units, which I can't fathom because I find the thing kind of dumb. The founder calls it "active intelligence reduction."

Maybe I just have dumb-thing aversion; I've never been able to use it. I visited Alibaba's Tongyi hardware exhibition recently — Fu Zai everywhere. The knockoffs were even dumber than the original. I thought I'd stumbled into a Fu Zai stupidity competition.

But JD's strategic investment team is still worth watching. Dada Nexus and AHS Recycle set the template: after investing, JD handed JD Daojia to Dada and Paipai secondhand to AHS. JD also made a move on Vivix AI's Liu Yu — maybe merchants will soon generate product pages in real time?

Oh right, nearly forgot another base-holding old friend: Baidu, one of the BAT trio. Anyone still remember? 🤓

Baidu Venture's investors plus entrepreneurs keep insisting the fund is independent, nothing to do with Baidu. But Baidu strategic investment folks have it rough. The bitterness, I feel it 😭

And Baidu-backed star founders — they'll grit their teeth and show up at conferences, but never voluntarily mention Baidu invested in them. Yes, Zhang Yueguang and Chen Mian, I'm looking at you.

But founders who took Baidu money really have nothing to complain about. After all, "every company Baidu invests in is a雪中送炭 [sending charcoal in snow] situation — they'd never take Baidu's money if they could get anyone else's."

All of the above is written for investors to see themselves in.

For founders, looking at it from the CVC perspective in reverse: this opportunity really is bigger than mobile internet. Back then, Meituan, Ele.me, JD.com, DiDi, Xiaohongshu, Bilibili — every mini-giant had to pick sides between Alibaba and Tencent, or play both.

Today's founders don't face this yet. Because in PC and overseas markets, Alibaba and Tencent don't have absolute dominance.

This phase is more like game development: genius creators don't need to worry about distribution yet, only whether their product is good enough. Unfortunately most people are just copying and hyping 😭

Or maybe the big reckoning just hasn't arrived — because the first AI application company to have all three BAT members as shareholders is (perhaps) about to emerge. Which company? Scroll back up.

(Article illustrations generated by ChatGPT; entirely human-written. Also: announcing our Saturday offline event.)