I'm here to congratulate LibTV.

"Love Mian TV Is Here"

Woke up this morning to the news that LibTV raised another $300 million... no, seriously, how is Mian pumping out this much gold? I'm actually crying 😭

Who said AI apps are under pressure? Who said the model layer dominates everything? Mian says he disagrees 😭

I wrote on Jike about my take on LibTV two months after launch. The gist:

LibTV captured market share through subsidies, aggressively adopted Seedance, brute-forced its way to miracles — genuinely putting ByteDance methodology into practice.

Subsidies inflate revenue, revenue attracts funding, funding accelerates subsidies — a three-way loop where the money on the books keeps growing, until money itself becomes the moat.

And that's exactly how the script played out.

Now with $300 million in the bank, I estimate LibTV has raised more than all its competitors combined. When other VCs look at the video space, they'd be hard-pressed not to factor Mian into their calculus.

If the guy had actually built from scratch himself, launched a few months later, missed the Seedance 2.0 traffic wave — maybe the video agent space would still be a fragmented battlefield of warlords. Not saying Mian is the Qin Shi Huang of it all.

Just goes to show: a market with real potential doesn't allow anyone to develop in quiet. If not ByteDance, there'd be wolfish companies like Lovart.

Not sure if it's because they took Tencent's money, but I'm already seeing LibTV ads constantly in my Moments feed. Next thing you know we'll be using LibTV in the nine-grid layout...

Feels like not only is Lovart essentially a Zhuanzhuan at its core, its ad spend is about to get as god-tier as Zhuanzhuan's too.

Announcing the fundraise during 618 — what's the implication? Are you guys doing the big sales festival too? Declaring that the next generation of "ten-billion subsidy" will be led by LibTV.

Still unclear how LibTV's unit economics actually work.

Some tell me they have gross margin, others say they're selling at a loss. All I can say is: if ARR as a metric can be gamed N different ways, then of course every methodology yields its own result.

Mian also disclosed that company ARR is approaching $300 million (I don't believe this number). Let me break down the current business:

Two cash cows: LibTV and Lovart;

The original white moonlight, Liblib;

The China version of Lovart, Xingliu;

And Zaoci, which already died — they were gearing up for the China version of Sora 2, didn't expect Sora 2 itself to surrender first.

At year-end they disclosed Lovart ARR was still $80 million, meaning most of the company's current ARR comes from LibTV. Yet LibTV's website traffic isn't particularly high.

So who are the real customers? From what I've gathered: AI short drama companies.

This makes perfect sense. Because AI short drama companies can bulk-harvest LibTV's 618 promotional subsidies. And short drama companies operate on simple logic: whoever's cheapest, I use. You try to price-war with LibTV, and Mian will show you his bank balance.

The outstanding founder Zhang Yueguang, analyzing the AI short drama ecosystem on Jike, demonstrated a level of insight rivaling an Alibaba P9:

"Many people may not realize that AI short drama is a magically unprecedented business model: first, video agent companies buy Seedance from ByteDance at high prices, then spend on traffic/KOL acquisition to get B2B customers (much of which also goes to ByteDance);

B2B short drama companies buy the tools, produce shows, distribute back to ByteDance platforms, then spend on ByteDance traffic (and this cost far exceeds production cost)"

Mian, formerly ByteDance 4-1, is the largest video agent company in this loop.

And so, LibTV, top AI short drama companies, and ByteDance form a perfect closed ecosystem.

Under discounts rivaling e-commerce's 618 mega-sale, Mian has turned himself into Volcano Engine's supersized KA (key account). At LibTV's current Seedance consumption volume on Volcano, no other ByteDance video agent department can even touch it.

In this dynamic, I believe LibTV should aggressively expand into investment. Invest in more video agents — then those video agents wouldn't need to buy Volcano quotas themselves, just use LibTV's. And nobody understands video agents better than LibTV.

By this logic, LibTV is the real Jinqiu Fund 👍

Clearly, having left ByteDance, Mian loves ByteDance more than ever — subsidizing his old employer with Tencent and Ant Group's money. How deeply affectionate.

Meanwhile, I'm genuinely curious about LibTV's real working conditions. Anyone want to write a In the LibTV — hit me up in the DMs.

At a time when AI tools are getting demolished by Claude Code, model routing platforms have become the interim answer for AI applications. Xianyu said when he chats with AI app founders, they're all praying for Manus and Mian to go public, to expand the application space. Can only say: Long live Saint Mian 😭

On such a fine day, Love Mian TV has nothing much to offer — got Brother Feng to record a video, sincerely congratulating him:

"Wishing the heir to the ByteDance way, the one who pumps Ant and Tencent gold to empower ByteDance, the流水圣体 [流水圣体], the only one who can battle world models and embodied intelligence — Chen Mian — may he keep pumping gold."

(Cover image generated by ChatGPT, purely human-written)

Subscribe to our Substack: funeralai.substack.com