Lovart Goes All-In on TapNow

"A Massive Zhuanzhuan" Lovart is a pretty good AI product. Not only does it let you use Nano banana🍌 on the cheap, but more importantly...

Alright, I can't keep this up. Lovart's main selling point really is just cheap Nano bananas. Its core function is basically a 🍌 relay station.

Yep, Lovart is essentially a Zhuanzhuan.

And hey, there's nothing wrong with that. Shells have their value. Among the sea of AI image-generation products, Lovart got its image editor out early and went hard on marketing blitz — there's definitely an AI-era Canva niche it could occupy.

Let's run the classic AI valuation math: does Lovart deliver 1-5% of Canva's user experience? Definitely.

Canva's valued at $42 billion. Lovart's currently over $1 billion, roughly 2.6% of Canva. Plenty of room to grow 🤓

Of course, our dear Lovart founder Mian Chen — always building, always pushing boundaries — hasn't stopped at Liblib, Lovart, Zaoci, and other products. He's now officially entered video generation.

Yesterday, I clicked on Lovart's new product LibTV with reverence. Landed on the homepage and thought I'd accidentally opened TapNow.

Quick primer on TapNow for everyone: it's a video workflow tool aimed at professional directors. Using it feels eerily like wrestling with Premiere back in school — steep learning curve. Not recommended for casual users who just want one-click video generation. But the control is granular; you can tweak individual details within individual shots of individual scenes. Great for Xinpianchang types doing commercial and TVC work.

One feature I love in TapNow is shot breakdown: upload a video, get a shot-by-shot script — super useful for derivative content.

LibTV conveniently supports this too. Very thoughtful:

TapNow's community has something called TapTV where you can browse others' work, clone their canvases, and study their approach, techniques, and prompts.

LibTV just so happens to have TV Show, which also supports shot breakdown. You can learn here too. How wonderful.

TapNow also lets you apply various edits to an image: repaint, erase, enhance, outpainting, cutout, multi-angle, relighting.

Wait, why did I say "also"? Could it be because LibTV does all this too? Indeed — repaint, erase, outpainting, cutout, multi-angle, relighting. Not a single one missing.

Click into the relighting details:

The full package.

The only difference might be that TapNow has "enhance" while LibTV has "HD" — but click in and it's basically the same thing.

Haha.

I thought this grid-splitting feature might be genuinely innovative.

But a friend pointed out that a tool called Aix Studio has grid splitting too.

Both also share a cleverly designed annotation UI:

Man, I don't even know what to say 😭

I'd suggest Mian Chen just put out a tutorial specifically explaining how LibTV differs from TapNow.

I'll start: first, the name is different. Second, LibTV has a Skill 🤓

Call it MCP, Skill, or CLI — doesn't matter. What it actually does is act as a connector, letting you send a message in Lark and have Lobster call LibTV to generate a video.

I predict Lovart will likely argue that their product is "the first professional video creation platform simultaneously serving humans and Agents," fundamentally different from TapNow which only serves human creators. Then they'll launch into grand narratives about how next-gen platforms are CLI-based and built for Agents.

But as one American journalist put it well: if something walks like a duck and quacks like a duck, it's a duck 🦆

I open LibTV and TapNow's websites and see similar UIs and interactions.

Then I suddenly remembered an article headline — Exclusive Interview with Lovart Founder Mian Chen: We Don't Have Product Managers, Only Designers

I strongly agree with this assessment.

Building products in the AI era really is undergoing a once-in-a-century transformation, because now designers can just study other products' interfaces and vibe code directly.

A PM friend compared the two products and sighed to me, "AI coding has advanced so fast — copying someone else's feature is way more efficient now," plus "TapNow doesn't have a 61% off membership prompt, so that's LibTV's original contribution."

Whoa, easy there — how can you say that? A new product's gotta go hard on promotions to get prices down for us users!

Writing this, I suddenly remembered that two months ago people were constantly discussing on Xiaohongshu whether to subscribe to TapNow or Lovart. I was baffled at the time — these are completely different things.

Currently unclear whether those posts played any role in Lovart's determination to "pay tribute."

Earlier still, another friend of mine, Julian — yeah, the guy who used to post 100 Jike updates a day.

When he saw TapNow last December, Julian immediately recognized its potential and told me on the spot he was going to build an open-source TapNow. Sadly, he was too lazy, and now someone's beaten him to the closed-source version 😭

I also asked the TapNow team for their take. They said, "Some interactions are getting major overhauls soon — the poorly designed parts got tributed too."

Why so passive-aggressive? Guess they're genuinely shaken.

The guy also shared a chat screenshot, slightly unhinged — a user who'd asked him for an invite code months ago suddenly messaged: "What's the relationship between Liblib's new LibTV and you guys? Did you strike some deal?"

He sat in silence for over ten minutes before replying, "Haha no deal, probably just tribute/imitation."

I could feel his sorrow through the screen 😭

Shells have their uses, shells have their moats. Last year a bunch of people cloned Manus, and Manus is still defining the category today. For creator tools, defining a product matters because it shows you've figured out where you're going next.

Equally important is the creator community — whether a tool can become a platform depends on how many creators are willing to build on it.

Sharing an AI film I recently watched.

Made with TapNow — you can still see @DiDi_OK's creative process on the site.

While I don't recommend TapNow for beginners, original vertical platforms like it have genuinely cultivated a crop of premium AI mid-to-long-form video creators.

I do wonder: if a tool excels at tribute, can creators trust it?

Running through Lovart's current businesses: domestically there's Xingliu, LibLib, LibTV; internationally there's Lovart and Zaoci. Five fronts simultaneously — is this the ByteDance 4-1 approach to startups, five horses galloping five ants 🐜

Still, Lovart as a product is decent. Designers may not have much loyalty, but throwing money around has genuinely built brand recognition. And honestly, small-town shop signage aesthetics still have massive room for improvement.

At least as a 🍌 relay station it's fine — a Zhuanzhuan business.

Keep spinning, keep turning. Spinning and leaping oh~

Anyway, Lovart is a massive Zhuanzhuan. It facilitates VC capital flow and accelerates user coin transfer. As for LibTV, I think its relationship to TapNow is like Zhuanzhuan to Xianyu.

So, in the AI era, is tribute the core competency? Is bold tribute a good thing? Will the best tribute artists win the most?

I don't know.

But I want to ask Mian: should the old Xingliu, LibLib, and Zaoci be listed for sale on Zhuanzhuan?

(Cover image generated by ChatGPT, purely human-written, always ready to crowdfund legal fees, and here's an outro 🎷)

🎷 Subscribe to our Substack 🎸

funeralai.substack.com