quietmoney

Lovart

An AI design Agent: users describe what they need in conversation and it generates a whole set of designs such as logos, posters, packaging, social media images, and video ads.

lovart.ai ↗AI creationChina 24.44%, United States 14.65%, Hong Kong 6.41%#74 of 606
LiblibAI (one of China's largest SD model communities) assembled a team 3 days after Manus blew up, copying Manus's 'Agent task flow + invite-code hunger marketing'. On 2025-05-13 it launched on X as the 'world's first design Agent'. Tech KOL reposts (interactions with Musk and others), invite codes resold for 500 yuan, and users spontaneously showing off a complete coffee-brand VI set that went viral filled a 100K Discord/waitlist in 5 days; it then moved to a subscription credits model, and the design drafts users produce keep bringing in newcomers on Xiaohongshu/X.One-line growth story, compiled by us from public sources; speculation is marked separately below.
Open Lovart ↗

Traffic model, last monthFrom first hearing about it to reaching the checkout page. Circles show how many passed through last month

Grey balls: target audience, not users yetGreen balls: Users (existing users who return to the canvas to keep designing, or people who keep browsing after arriving)One ball ≈ 266K people (100 balls every 5 seconds stand for one month)Click a circle for details and source links
Where the numbers in this chart come from

The main engine is viral spread: direct visits are 63%, the search terms are all branded terms like lovart and loveart, and full design sets that users post, plus friend recommendations, bring people in.

The launch was a turbo boost: in 2025 it copied Manus's invite-code playbook and launched on X, with about 20,000 people queuing on day one, invite codes resold for 500 yuan, and a 100,000-person waitlist in 5 days; checkout page visits have now fallen from the peak.

By our estimate, about 40% of visits are existing users returning to the canvas to keep designing; about 1.5 in every 100 visits click through to the checkout page.

Model: Lenny's racecar growth framework. Engine = viral spread (users post branded design drafts and recommend it to friends, bringing people to search the brand name or open it directly); turbo boost = the 2025 invite-code launch and KOL reposts; lubricant = projects on the infinite canvas and the credit subscription, which bring existing users back to keep working.

Real data: 2.3M visits, -13.82% month over month, largest channel direct at 62.71%, top three channels ranked (direct, organic search, organic social), organic search 77.86% within search and paid search 22.14%, 5,700 search terms, top search terms (lovart, lovart ai, loveart, loveart ai, love art), top referrals (saaspo.com 4.38%, lovart.io 3.49%, sonicourses.com 2.96%), social sources (YouTube, Pinterest, Instagram), countries (China 24.44%, US 14.65%, Hong Kong 6.41%, Taiwan 6.1%, Brazil 5.01%), bounce rate 34.13%, 4.06 pages per visit, average stay 3 minutes 52 seconds, all from the Similarweb public page, last month; the public page does not give the shares of the second and third channels. 34,384 visits (about 34,400) to the Stripe checkout page come from the quietmoney checkout page data, 27,237 in the month before, 29,729 in the month before that, and an average of about 44,100 per month in the months earlier.

Supporting evidence: Incubated by LiblibAI; CEO Chen Mian is the former head of global commercialization at Jianying/CapCut; after Manus launched in 2025 the team sprinted for a month and launched on X by invitation as "the world's first design Agent", with about 20,000 people queuing on day one and invite codes resold for 500 yuan (36Kr); a full coffee chain brand design set posted by a user got 1M+ views, the waitlist hit 100,000 in 5 days and Discord passed 100,000 in 5 days (TechCrunch sponsored post, company figures); first-month web visits were about 300,000; at the official launch it claimed 800,000 users in 70 countries; knockoffs that do not subscribe appeared on HN; Lovart 2.0 was released in 2026.

Inference: The two channel circles are real; how each route is split is an inference: direct 1,442,300 = existing users 800,000 + posted designs 300,000 + friend recommendations 250,000 + tool directories and third-party sites 92,300; other channels 857,700 = posted designs 350,000 + tool directories and third-party sites 207,700 + existing users searching the brand name 150,000 + friend recommendations 150,000. Semrush desktop data shows direct at about 93.1% and search at about 6.9%, which suggests existing users on desktop almost all open it directly, and social and search come more from phones. That 3% of people who see a posted design click through, 40% of people recommended by friends visit the site, and 10% of people who see it in a tool directory click through are guesses. 22.14% of search is paid search, which suggests it also buys search ads, at small volume, and is not drawn as a separate route.

Checkout visits, last month
34,384
#74 of 606
vs prior month
▲ 26%
27,237 → 34,384, gain +7,147
Website visits, last month
2,300,000
Similarweb, last month; about 1.5 of every 100 visits reach checkout
Similarweb global rank
22,332
#251 in Finance/Investing
Take-off year
2025
How it startedNew-platform windowGrabbed the window when a new platform or model appeared
New-user source nowOtherMostly referrals, email or paid, or mostly direct with no clear source
Growth engineCreators or communityCreators keep bringing users, or the maker runs a community
Spreads purely by itselfMixed

Checkout visits over time

44K4–6 months ago avg
30K3 months ago
27K2 months ago
34KLast month

Checkout: checkout.stripe.com

1. Productwhat it sells, to whom, and how it charges

Product
An AI design Agent: users describe what they need in conversation and it generates a whole set of designs such as logos, posters, packaging, social media images, and video ads.
Users
Small businesses, marketers, and freelance designers who need design drafts fast (many in Chinese-speaking regions and the US)
Positioning
Calls itself the first AI design Agent, producing a full set of brand materials from a single conversation
Pricing
Free to start; Basic/Pro/Ultimate/Team subscriptions, monthly or annual, plus top-up credits you can buy; exact price not found Billing: subscription Free tier: yes
Price source
lovart.ai
“Title "Lovart – World's First AI Design Agent | Create Pro Designs in Minutes"; sample prompts for menu and wedding posters”Quote from the site

2. Channelswhere new users come from and what the traffic looks like

New-user source: Other (Direct / referral / email / paid). Direct 93.1%, search 6.9%; search terms are all brand keywords (lovart, loveart; brand keywords 100%), so new users come from brand buzz. Referrals come from saaspo.com 4.38%, lovart.io 3.49%, and sonicourses.com 2.96%, and social is YouTube, Pinterest, and Instagram; returning users go back to the project page to keep creating.

Last month's traffic sources

Search 7%Direct 93%

Semrush source breakdown (desktop)

Five-class attribution

Attributable new users, last monthShare
Other100%

Direct + brand search 100%. Direct and brand search count as viral only when the site shows sharing mechanics; the largest share wins, under 5% means Other.

Countries

CountryShare of visits
China24.44%
United States14.65%
Hong Kong6.41%
Taiwan6.1%
Brazil5.01%

Countries (Semrush): Bangladesh 23.0%, China 21.7%, United States 12.2%, Japan 6.7%, Brazil 5.1%

Language

English

Social sources

YouTube, Pinterest, Instagram

AI assistants

SourceShare
ChatGPT0.12
Gemini0.04
pruna.ai0.03

AI referrals (Semrush): ChatGPT 7,688 (0.12%), Gemini 2,566 (0.04%), pruna.ai 1,835 (0.03%), lovarta.ai 949 (0.01%)

Keywords

5.7K keywords · brand ~100% · organic 77.86% / paid 22.14%

Similarweb top 5 keywords
lovart
lovart ai
loveart
loveart ai
love art

Referring sites

30 referring sites

SiteShare of referrals
saaspo.com4.38%
lovart.io3.49%
sonicourses.com2.96%

Referrer categories

CategoryShare of referrals
Ecommerce and Shopping - Other25.84%
Search Engines12.16%
Computers Electronics and Technology - Other9.08%
Graphics Multimedia and Web Design6.75%
Programming and Developer Software6.24%

Site metrics (Similarweb, last month)

Visits, last 3 months
2.3M (3 months)
Visits MoM
-13.82%
Country rank
China #3161
Category rank
Finance/Investing #251
Bounce rate
34.13%
Pages / visit
4.06
Visit duration
00:03:52
Industry
Investing
Top 3 channels
Direct › Organic Search › Organic Social (top channel 62.71%)
Semrush monthly visits
3 months ago 5,470,929 · 2 months ago 5,310,549 · Last month 6,173,400 Semrush scale, ~3–4× Similarweb
Semrush organic / paid search visits
449,845 / 36,762
Semrush source breakdown (desktop)
Direct 83.6%, google.com · Google organic 5.5%, google.com.hk · Google organic 0.7%, accounts.google.com 0.7%

Growth-path traffic modelHow many people passed each step from launch to take-off

🔒 Unlock with quietmoney ProSee how many people passed each step from launch to take-off

3. Growth pathhow the first users arrived and what kept it growing

Classification

Basics

Timeline

Growth loop

Revenue and scale clues

Confidence and gaps

🔒 Unlock with quietmoney ProSee the timeline, growth loop, revenue clues and every source

Other sites that started via “New-platform window”

Similar sitessame category, also growing via Other

FAQ

How many checkout visits does Lovart get per month?

About 34,384 last month (Similarweb estimate), #74 of 606; the month before was 27,237. Opening checkout is not paying, so this is not revenue.

Where do Lovart's new users come from?

New-user source: Other (Direct / referral / email / paid). Mostly referrals, email or paid, or mostly direct with no clear source Direct 93.1%, search 6.9%; search terms are all brand keywords (lovart, loveart; brand keywords 100%), so new users come from brand buzz. Referrals come from saaspo.com 4.38%, lovart.io 3.49%, and sonicourses.com 2.96%, and social is YouTube, Pinterest, and Instagram; returning users go back to the project page to keep creating.

How much does Lovart cost?

Free to start; Basic/Pro/Ultimate/Team subscriptions, monthly or annual, plus top-up credits you can buy; exact price not found

Dossier confidence: medium. Parts 1 and 2 come from reading the site and Similarweb data; part 3 from public reporting, founder posts, Wayback snapshots, WHOIS and similar, with speculation marked.

The number: checkout visits is Similarweb's estimate of visits from the site to a checkout payment page; opening checkout is not paying, so this is not revenue. Small sites land on fixed buckets.

Data through last month. Page updated 2026-10-31.