Agnes: A Company That Only Copies and Hypes Its Way Into AI Circles

"The Most Shamelessly Hyped AI Company of 2025"

I never would've paid any attention to Agnes AI if I hadn't come across this tweet: a $100 million valuation this round, $300–500 million next? This sounds like an absolute AI star — how come I've never heard of it? And this fundraising language, why does it reek of scam?

Anyone who's raised money knows that "financing is in the close stage" and "has received term sheets from multiple prestigious institutions, several of which have completed due diligence" are two statements that don't belong together. Because receiving a term sheet is still a long way from closing a round — and they've been touting their $100 million valuation since August.

I've written about so many AI fundraises, and this is the first time I've seen a company brag in its PR about receiving lots of term sheets. Then the next line: "preparing its next round, with an expected market valuation of $300–500 million." This is pure bait — telling investors, better get in now or I'll quintuple on you. It's basically got "trying to scam money" written across its face.

With that kind of nerve, I got genuinely curious about the product. I open the page: a completely generic chat box. A toolbar with various productivity tools, zero innovation — I immediately see visions of Quark and GenSpark.

I asked it to generate a PPT comparing Flowith and Agnes. Utterly mediocre, worse than what I could make for free with Quark. And the content is so thin — why does it say Agnes has an "AI self-developed large language model"? What does "AI self-developed" even mean?

So when the official line says they're building "agents for the common people," does that mean common people can just throw together any old agent and make do? They also want to "truly bring the 99.5% of users who've been isolated and ignored into the world of AI." If ordinary users' first AI experience is this, they're truly screwed — they'll think all AI is garbage.

For comparison, I had GenSpark make the same thing, and it turns out Agnes is a pixel-for-pixel copy. Left sidebar: views, code, canvas. Right sidebar: editing. But the copy is so bad that calling it a budget GenSpark is almost insulting to Kun Jing. And this website — pure vibe coding. Did they spend all their money on traffic and PR, with nothing left for frontend devs or designers?

My advice: regenerate this with Gemini 3.0, it'd probably be better. I also looked into the company and founder's background. The founder is Yang Tong. In 2019, he started Kaiwei Technology in Nanjing. What does this company do?

— Official description: "Kaiwei Technology focuses on the 'last meter' critical scenarios in retail, building an omni-channel reach path covering consumers' complete shopping cycle 'in-store + online + off-site,' providing brands with data-driven, multi-dimensional marketing solutions." Then Agent hype took off in the first half of this year, Yang Tong became Bruce Yang, the Nanjing company relocated to Singapore for hype, and a SaaS company serving offline retail transformed into "an innovative enterprise focused on enterprise-grade AI Agent platforms."

Glorious Singapore, glorious era of hype — giving SaaS companies that couldn't make it a chance to start fresh and re-hype themselves. So my suggestion for investors meeting them: ask if they're actually in Singapore. Because from what I see, Bruce Yang's WeChat Channels IP location is basically always Jiangsu.

Agnes also launched on Product Hunt in July. Today they claim over 200,000 DAU, with 3 million new registered users in two months since launch. Who's actually using this? I checked their traffic sources — top five: Philippines, India, Vietnam, Indonesia, and Brazil. All Southeast Asian brothers.

I get it now. The man's Singapore's Robin Hood, scamming US dollars to subsidize Southeast Asian and Latin American commoners, redistributing wealth to promote shared prosperity for all the world's people — truly a heart of gold. He told investors he's mainly targeting Android-dominant markets with engineering optimization. So Agnes is positioned as an Android AI app, and he swore up and down they haven't paid for user acquisition, mainly driven by KOLs.

Are Southeast Asian KOLs really that powerful? Here's a simple truth: a normal company or product has traces of time. There's no such thing as an overnight viral niche AI app — only bought Southeast Asian users.

This February, when Kaiwei Technology launched its new product, Bruce Yang said they should work hard to think about how to become the next DeepSeek. A few months later, the man figured it out: endless clout-chasing. Dude's constantly attending forums, giving interviews saying he wants to be the next DeepSeek — sometimes Jiangsu DeepSeek, sometimes Singapore DeepSeek, later saying he's most likely to become "Southeast Asia's DeepSeek."

Might as well just say he is the real DeepSeek. The day after Manus launched, the Kaiwei Technology WeChat account posted: "Manus's Features Were Already Implemented on Roles Last Month?"

Their latest post now: "The DeepSeek Moment for Agents Is Here! This Dark Horse Is Using a General-Purpose Foundation to Mass-Produce China's Manus." One post, two bandwagons.

Then they started riding GenSpark's coattails. Lovart raised money, they rode that too. Recently they're chasing Instagram and Snapchat. Let me catalog Agnes's current labels: Nanjing DeepSeek, Jiangsu DeepSeek, Southeast Asia DeepSeek, Singapore DeepSeek (this retreat route eerily resembles a certain bald historical figure 🧑‍🦲);

Ending the Office era; Caught up to GenSpark in 3 months, closing in on Manus; Traffic already approaching design-category leader Lovart;

Breaking Instagram and Snapchat growth records. So many labels — keep clout-chasing and hyping until you become AI's own Big Bing. I predict the man's next target is Liu Yu's $1.3 billion multimodal AI play. Perfect, Liu Yu doesn't even have a product yet — he can chase it into the void.

This cracks me up. I genuinely love this company. I hereby declare Agnes the most shamelessly hyped AI company in China for 2025. The man has his own WeChat Channels too — when you have time, could you recite everything above? If you can keep a straight face, I'll believe you can end the Office era.

I also asked some investor friends. Most of them have actually talked to the company, and someone said they were quoted $300 million, plus they're preparing to launch AI headphones. I was stunned, lamenting to a friend: these days an AI company can copy something identical and still scam money? He asked me back — and he got me 😭 inspired by Morris.

)