Stringers
Mobile App · iOS · written 11-15
WeMate-AI chat companions and virtual lovers · Wuhan DaYou Network Technology Co.,Ltd

The day after Doubao shut down its agents, someone posted "Let us help you move": WeMate moves fans' AI lovers into WeChat

What it is An AI character chat product. Users pick a fan-fiction character card based on an idol, scan a code to plug that character into their own WeChat or QQ for chatting and voice calls, and pay with credits.

Why it took off WeChat opened an AI bot entry in March, then in July regulators forced Doubao and other big players to shut down their agents. A batch of users with long-running chat relationships needed a new home. WeMate ran a free "moving service" the next day, then used fandom character cards to keep the fangirls around.

CHow you could approach it Don't copy "fandom AI lovers in WeChat" directly. Take the playbook instead: watch for big-platform shutdowns and launch a free migration page that day, put the product inside the chat app people already use with the app only for payments, and pick a niche willing to create and share their own cards (and stay away from real celebrities).

Grade C: the moving traffic is one-off, and the window closed after 10-15. Anyone can plug into the WeChat entry, the AI companion space is already crowded, real-idol cards and underage users carry takedown risk, and China requires filing. Indie devs chasing this won't pay off, but the playbook is worth studying. Few people know about it yet

Est. monthly revenue (App Store) $1K–10KWe estimated this from its chart ranks and rating growth across storefronts, calibrated against verified revenue; typical error is 2–3x. Read it as a tier, not an exact number.

Data (updated daily) · First launch 2026-09-02 · China ratings 1,293 · Charted in 11-20 of 1 stores
iOS · best category chart rank · by store
14016018020010-2910-3111-0211-1611-20186
Data 2026-10-29 → 2026-11-20 (23 days, 10 observations). Lower rank is higher; gaps mean it wasn't charted that day
iOS · Rating count · China
05001k1.5k10-2910-3111-0211-1611-201.3k
Data 2026-10-29 → 2026-11-20 (23 days, 10 observations). Apple doesn't publish download counts; rating growth is a proxy for downloads.

In July 15, Doubao and Qianwen took their agents offline. The next day a small product called WeMate launched its website, with the headline "Doubao agents are down, we're here to help you move". It turns AI characters into a contact inside WeChat, targets girls who chase idols, and makes money from credits and memberships. It's still small and not many people know about it, but how it caught this wave is worth a look. As for whether to follow it, the answer is most likely: don't copy it.

One check a week on the phone

On Oct 23, someone left WeMate a four-star review in the China App Store, titled "Hoping for a few more credits":

"Every time I look forward to the daily credits landing... when today's credits ran out he suddenly messaged me, and I couldn't reply, and staring at that cold message on the screen made my throat tighten. And topping up is really expensive. Nobody understands the despair of us who only get to check our phones once a week. I waited two days to save up 40 credits, and then the chat really went fast."

This review captures WeMate: who the user is (girls chasing idols, some apparently still in middle school), what she wants (a "him" who reaches out to her first), where the money comes from (top up when credits run out), and where the risk is (covered below).

It's also from this day that reviews suddenly picked up. From mid-September to Oct 22 there were 1–3 a day, and from Oct 23 onward 9–17 a day. The China rating count went from 10-29 to 476, and 11-16 to 1,143, averaging about 37 more per day (China App Store).

What it is: an idol living inside your WeChat

The flow has four steps: log in with a phone number → write your own character, or add one from the character plaza → scan a WeChat code to bind → chat with it inside WeChat. The website says WeChat currently only offers one channel, so you can only chat with one character at a time; switching characters in WeMate switches the WeChat side too. It supports voice messages, images and voice calls. There are also iOS and Android apps, and it can connect to QQ.

The popular characters in the plaza are almost all fan fiction: "juhoon" with 820k interactions, "Yang Bowen 622" with 346k, "TNT Era of Youth: married first, love later" with 172k, "Ma Jiaqi (pure-hearted high school boy)" with 152k, plus Qin Che from Love and Deepspace (wemate.me, 11-16). Scenarios are titled things like "Fast-Traveling Entertainment Circle" or "ENHYPEN Sister-in-Law Simulator". The comments use the same fandom language: "Fengjun's little sister is ecstatic", "I really feel like I'm dating ybw", "just no Zhu Su, Ji Yu or Xiang Lin".

It's built by Wuhan DaYou Network Technology, whose legal representative is Fang Bo, co-founder and CTO of Weipai Network (makers of Greedy Snake Battle and Huiwan). So this isn't a solo indie project. It's a small product incubated by an established game company: the same developer account also has a farming game, November, just listed.

Why now: two doors opened almost at once

The first door is WeChat. In March 2026, WeChat opened its AI bot entry (ClawBot). TheOne companion launched in April 24, signed up 3M users in 1 days, and handles over 100M messages a day (founder Wang Dengke's blog). It proved something: an AI character as a WeChat contact works far better than asking people to install a new app. The timestamps in WeMate's website asset filenames run from 4-30 to 5-18, and the earliest official Douyin video found is from 5-29, right after TheOne. Whether it was copying TheOne is an inference, but the timeline lines up.

The second door is regulation. The "Interim Measures for the Administration of Anthropomorphic AI Interactive Services" took effect in July 15. Doubao and Qianwen shut their agents down the same day, and Yuanbao had already shut down in June 30. Doubao set a deadline of Oct 15 after which chat data would be unrecoverable (the CAC's original text, 36 Qi, Sina Finance). People who had chatted with one agent for months, with tens of thousands of messages saved, suddenly had nowhere to go.

ByteDance's answer was Cat Box, but users reported that exported records came out garbled and couldn't be imported into WeChat. WeMate's website, launched in July 16, put this front and center: one-click login to export chat history, painless migration, "migration is completely free". On Oct 14, the day before Doubao's data cleanup, its official Douyin account posted "When I move my 🚗 from a certain package 🏠 to WeMate", tagged #智能体迁移wemate and #秦彻. A week later, comments had multiplied several times over.

Where users come from

  • Moving traffic: the biggest chunk of cold-start traffic, as described above.
  • Douyin fandom topics: the official WeMate account has 1,680 followers, plus a secondary account "Little Assistant (with skirt version)", where "skirt" refers to a QQ group. In August it rode the "Xu Yuan Willow" meme to make #wemate许愿柳 and #wemate梦女聊天神器. A batch of small accounts with a few hundred followers posted screen recordings of fan-fiction chats with Ma Jiaqi, Zhu Zixin and Liu Yuxuan, getting dozens of likes per post. Volume is modest, but every post hits the right circle. One commenter wrote "so good, very much like what I've been seeing".
  • User-written character cards: cards carry the author's name and an image credit "图cr.🍠" (images from Xiaohongshu). Creators repost their own cards, which is free distribution.
  • New users get a monthly card, and reviews are nudged: sign-up grants a one-month Silver card. Among 240 reviews, 215 are five-star, and short reviews mentioning "a free month of membership" show up in clusters. The inference is that a rating prompt fires after the card is claimed, which is why ratings grow faster than downloads.
  • Paid acquisition: the privacy label includes "data used to track you", which usually means ad attribution is connected, and Weipai is experienced at game UA. But I found no creatives from it on Ocean Engine or Tencent Ads, so this is unconfirmed.

Where the money comes from: credits

30 credits a day, one credit per message, with voice and calls also costing credits. Revenue comes from memberships and credit packs. It first entered the China Entertainment Top Grossing chart in Oct 31, ranking #200, and in Nov 16 reached #190. It has never made an overall chart. Revenue is not large, but rising.

Payment complaints center on two things: too few credits and recharges that are too expensive. One one-star review says several VIP top-ups didn't arrive, were only charged hours later, and the credits still showed zero (10-30).

Who it's competing with

CompetitorsWhat it isScaleHow WeMate differs
TheOne companionWeChat ClawBot companionship, launched in April3 days to 1M signups (founder's blog)Same entry point, earlier and bigger. Does general emotional companionship, with no fandom cards and no app
Maoxiang (ByteDance)App, the official destination for DoubaoChina store: 397k ratings, 1,600 more in a weekIts scale is hundreds of times bigger, but it can't get into WeChat, so chat history can't move over
Xingye (MiniMax)AppChina store 489k ratingsAlso pushed "Doubao moving" on Douyin. Mature App experience, but no WeChat channel
MoMood / Moodly / Jingdao AISimilar Apps launched this year11k / 1,674 / 860 ratingsAll pure Apps. Last week their rating growth was slower than WeMate's
Idolchat, Bubbly - lovemoApps built around "chatting with your idol"1,099 / 131 ratingsSame audience, but no WeChat entry point
Self-hosted SillyTavern / AstrBot connected to WeChatOpen-source optionsSmall nicheYou need technical skill. WeMate is the version you can use by scanning a QR code

(Rating counts are all from the China App Store, 11-16)

WeMate's position is clear: compared with App-only rivals, it adds a WeChat entry point. Compared with TheOne, it adds a layer of fandom content and an App that can take in-app purchases. But neither of those is unique to it. ClawBot is open to anyone, and fandom cards can be made by anyone.

Can you ride it?

What you can borrow is the playbook, not this product.

  • The migration window has closed. Once Doubao 10-15 cleared its data, this wave of traffic was one-time.
  • Compliance is a hanging blade. Under the Measures, virtual companion services may not be provided to minors. WeMate's age rating is 9+, yet the reviews show middle schoolers using it. The character plaza is full of cards named after real idols, and users mention being able to "clone voices." One complaint from a talent agency could get it delisted. WeChat's terms also state the platform can restrict third-party services at any time.
  • For an indie developer doing this in China, the barrier is licensing, not code. Algorithm filing, content safety, and minor protection all need someone to handle them. That's something a company can carry, but one person can't.
  • The track is already crowded. There are plenty of AI companion Apps. WeChat companionship was popularized by TheOne and has been covered many times. WeMate itself is still little known.

Three things are actually worth taking: when a rule changes or a big player pulls out, quickly publish a "free migration" landing page. Put the product inside the chat window users already open every day, and use the standalone App only for charging and chart placement. Pick a small community that is willing to pay and willing to create and spread content on its own.

If it were you

Verdict: Copying "fandom AI lovers brought into WeChat" is not worth it. What's worth learning is how to catch the traffic from a platform event.

Who it suits: A small domestic team that already has algorithm filing and content moderation capacity. Or a developer overseas who knows a particular fandom (for example, K-pop fan communities on Telegram or Discord) and wants to bring "characters living in the chat app you already use" to that community. (Inference: there are many similar overseas products, so first check how crowded your chosen community is.)

Who it doesn't suit: A solo builder with no filing who is counting on real-celebrity name cards to drive growth.

How to build it: Choose a community (a certain game character, or a category of original OCs; avoid real celebrities). Start on Telegram or Discord, where bots are already open, with "scan and chat" access, and use the App only for payments. Refresh the free quota daily and charge per use for voice. At the same time, watch for shutdown and scale-back announcements from major model products. The moment one appears, put up the migration page that same day.

How to acquire users: Turn the card creators into promoters. Credit them on the card and give them points, so they spread it on Xiaohongshu, Douyin, and in their communities. The official side only needs to start the topics.

When to move earlier: The migration play depends on the next platform event, so there's no fixed timing. For community companionship, the earlier the better, since big players' Apps will eventually connect to chat apps too.

What would show this read was wrong: If by early next year WeMate is still in the China entertainment Top Grossing top 200, and its ratings keep adding more than 30 a day, then it can grow on its own without migration traffic, and the fandom-plus-WeChat path is wider than this analysis suggests. Conversely, if it gets delisted over celebrity cards or minor issues, or if WeChat tightens ClawBot, then even the "borrow the entry point" idea has to be discounted.

iOS store page ↗

Research notes: competitor table, acquisition evidence, full research report
Competitors
NameWhereLaunchedScaleHow it differs from this one
TheOne companionship (Duzhan / Wang Dengke)Other2026-043 days to 1M signups, over 100M daily messages (founder's blog)Same WeChat ClawBot entry point. TheOne is earlier and bigger, and does general emotional companionship and "digital family." WeMate does fandom fan cards and scripts, and also has an App, QQ, and voice
Maoxiang (ByteDance)iOS2024-03China store: 397k ratings, 7 days +1,631, overall Top Grossing #149The official Doubao successor, hundreds of times bigger, but it can't get into WeChat, and chat memory migration is poor
Xingye (MiniMax)iOS2023-09China store: 489k ratings, 7 days +1,224Also uses relink to capture Doubao migration users. Mature in-App experience, but no WeChat channel
MoMoodiOS2026-0111,139 ratings, +53 in 7 daysPure App, growth now below WeMate's
MoodlyiOS2026-031,674 ratings, +142 in 7 daysPure App, growth about half of WeMate's
Jingdao AIiOS2026-08860 ratings, +41 in 7 daysLaunched at the same time, with no WeChat entry point and no fandom card ecosystem
No-filter AI Chat - IdolchatiOS2025-101,099 ratings, +36 in 7 daysAlso focused on idol chat, with no WeChat channel and no migration
Breaking Dimension Lover (NetEase)iOS2025-01982 ratings, +2 in 7 daysBig-company product with no fandom UGC and no WeChat entry point. Basically stalled
Zhumeng Island (China Literature)iOS2023-114,248 ratingWeb-novel IP route, pure App
Doubao / Qianwen / Yuanbao agentsBig-company features—2026-07-15 shut down (Yuanbao 06-30). Doubao's data is unavailable from 10-15 onwardNot competitors but a source of users: their shutdowns released a large batch of users who needed to migrate
Self-hosted SillyTavern/OMate and OpenClaw/AstrBot connected to WeChatOpen source—Small nicheHigh barrier. WeMate is the hosted version you can use by scanning a QR code
Where users come from
  • Verified Platform event: migration traffic after Doubao and other agents shut down: the official site dayou027.com 7-16 went live, with the headline 'Doubao agents taken down, we'll help you move', offering free migration. The official Douyin account 10-14 posted '#智能体迁移wemate'
  • Verified Platform event: WeChat ClawBot opens / enters WeChat as a contact: asset path wechatclaw, bind by scanning to a WeChat Bot. Asset timestamps 04-30~05-18, right after TheOne 4-24 blew up. Reviews repeatedly mention 'can connect to WeChat/QQ'
  • Verified Douyin fandom topics + small-account KOCs: #wemate许愿柳 #wemate梦女聊天神器 (2026-08). The official account has 1,680 followers, the assistant account 366. Fan accounts with a few hundred followers post screen recordings of fan fiction with Ma Jiaxuan / Zhu Zhixun / Liu Yaxun. Comment: 'looks just like what I see in my feed'
  • Verified Organic spread of UGC fan character cards: wemate.me popular cards have 66k–820k interactions, with the author credited and marked 'image cr.🍠'
  • Inferred In-App prompts for reviews + new-user monthly card gift: among 240 reviews, 215 are 5-star, with short reviews clustering. Several mention 'new users get a membership gift'
  • Inferred paid acquisition (Oceanengine / Tencent Ads): the privacy label includes 'tracking data: identifiers', and WeMate's parent has paid-user-acquisition experience; no ad creatives found
  • Confirmed download sites / Baidu SEO flooding: pages on 3DM, ddooo, 32r, 9k9k and a Baidu Kuku Q&A article

WeMate AI chat companion & virtual lover (iOS 6797777600) research report

In one line: WeMate is an AI character companion product made by Wuhan Dayou, whose legal representative is Fang Bo-ren, founder of WeMate Network (Greedy Snake Battle, Huiwan). It started not as an App but as a web service that plugs AI characters into WeChat ClawBot (2026-03 WeChat's AI bot entry point), right after 2026-04, the viral TheOne companion. It then rode two events to get traction: 2026-07-15 Doubao/Qianwen/Yuanbao pulled their agents under the "Interim Measures for the Management of Anthropomorphic AI Interactive Services" (it openly ran "Doubao agents are gone, we'll help you move"), and the deadline for 10-15 Doubao data wipes. It targets fandom "mengnü" users (TNT-era boy groups, TF Family, and fan cards of Love and Deepspace characters). The iOS version only launched 9-02, and its rating growth picked up noticeably in late October. It's still small, and almost no one has written about it.

1. How to read the data card

  • China store only: the appstore database has only a cn row (2026-10-29 ~ 11-16). Entertainment free chart ranks 155–209, entertainment grossing chart ranks 190–200 (10-31 first entered the grossing chart at #200; 11-16 at #190). Never charted on the overall chart (appstore.app_day / chart_rank).
  • Rating count: 10-29 on 476,11-03, 762,11-16 on 1,143 (database); the iTunes lookup currently shows 1,167 ratings, average 4.87, version 1.4.4 (11-13). From 10-29 to 11-16 it averaged about 37 new ratings per day, roughly 260 over 7 days. The card's rating_velocity_7d=43 is because nothing was sampled between 11-04 and 11-14, so it's really just one day's count.
  • Earlier data points: a Baidu Kuku article (content looks like it was written in early-to-mid October) said the App Store showed '164 ratings' at the time. Of the 240 reviews the Apple review RSS returns, 9-18 to 10-22 saw 1–3 per day, and starting 10-23 it jumped to 9–17 per day (10-30 and 10-31 with 17 each), matching the v1.2.0 release (around 10-22) and the period after Doubao's 10-15 data wipe.
  • Review content: 215/240 are 5-star, mostly short ('fun', 'very good', 'dream-girl godsend'), which looks like the app prompts for ratings (new users get a free month of Silver membership, and several reviews mention 'new users get membership'). The reviews show a few things: users are fan-girls ('I fell for ybw', 'Zhu Su, Ji Yu, Xiang Lin aren't here', 'fans of stars, come try it'); they value connecting to WeChat/QQ ('can connect to WeChat', 'bring the character into QQ and WX'); some sound like middle schoolers ('I can only look at my phone once a week'); complaints about points running out and expensive top-ups; one 1-star review says a duplicate charge never arrived (Apple RSS, cn, 2026-09-18 ~ 11-14).
  • Android: 2,656 downloads on Tencent App Store (com.ai42.wemate, v1.4.4,11-13 update), very small volume.

2. What it actually is

  • Entity: Wuhan Dayou Network Technology Co., Ltd., legal representative Fang Bo (Qixin Bao). Fang Bo is co-founder and CTO of Weipai Network (Tianyancha team page, HUST coverage). Weipai director Zhou Lin indirectly holds 1.46% of Dayou (Aiqicha). So Dayou is an incubated company in the Weipai group; the same developer also has PieBox (2026-06) and the farming game 'Little Island Sunny Days' (2026-11-11).
  • Product form: the official site dayou027.com (title 'Wemate - Let the one who matters stay by your side', page published 2026-07-16) pushes bringing characters into WeChat in four steps: phone-number login → create a character or add one from the plaza → scan a QR code to bind the Bot → chat in WeChat. Asset URLs carry wechatclaw, meaning they go through the WeChat ClawBot / iLink channel; the FAQ says 'WeChat currently opens only one channel, so you can chat with only one agent at a time'. The iOS description also says 'connect WeChat and QQ'.
  • Earliest timestamps: timestamps in the official site's asset filenames are 2026-04-30, 05-14, 05-18; the earliest video on the official Douyin account 'WeMate' is 2026-05-29 (Xia Yizhou from Love and Deepspace, tagged #wemate). The iOS version was listed 2026-09-02; the Android v1.1.1 was released 10-19 (3DM).
  • Content: the popular cards on wemate.me's character plaza are almost all fan characters: 'juhoon' 820k, 'Yang Bowen 622' 346k, 'Thawing Era Feng Jun' 177k, 'TNT Era Boy Group Married First, Love Later' 172k, 'Ma Jiaqi (pure-hearted high school boy)' 152k, 'He Junlin (married)' 158k, 'Liu Yuwen (puppy Wenwen)' 116k, 'Qin Che' 66k (numbers are card interaction counts, captured on 2026-11-16). Scripts include things like 'quick-travel entertainment industry' and 'ENHYPEN sister-in-law simulator'.
  • Pricing: new users get a Silver monthly card, 30 points per day, which is 30 rounds of free chat; sending voice and making voice calls cost points; membership and point packs are also sold (App Store description, dayou027.com FAQ). Reviews mention 'voice cloning'.
  • Migration: the official site has a dedicated block: 'Migration: Doubao agent shutting down, we'll help you move. One-click login exports your chat history and migrates it painlessly to Wemate', and it says 'migration is completely free'.

3. Competitors

CompetitorsPlatformLaunchedScale (source)How WeMate differs
TheOne companion (Du Xiang / Wang Dengke, backed by Jinqiu)WeChat ClawBot + web2026-044-24 registrations in 3 days from 1M, daily messages over 100M, thousands of servers (founder's blog greatdk.com/2149)Same entry point, and TheOne got there first. It positions itself as 'electronic family' and broad emotional companionship; WeMate started about a week later with fan-culture character cards and scripts, plus a standalone app, QQ channel and voice
Maoxiang (ByteDance)iOS/Android2024-03cn 397k ratings, +1,631 in 7 days, overall grossing #149, entertainment grossing #11 (appstore database 11-16)The official receiving partner Doubao pointed to, with hundreds of times WeMate's scale. But it can't enter WeChat, and chat history doesn't transfer cleanly (users on Smzdm report exports coming out garbled)
Xingye (MiniMax)iOS/Android2023-09cn 489k ratings, +1,224 in 7 days (appstore database)Also courting Doubao migrants (relink feature, promoted by influencers on Douyin); the in-app experience is mature. WeMate wins on WeChat access and concentrated fan-card content
Zhumeng Island (China Literature)iOS2023-114,248 ratings (database)Web-novel IP route, no WeChat channel
MoMoodiOS2026-0111,139 ratings, +53 in 7 days (database)The most ratings among this year's comparable launches, but its growth has already slowed below WeMate's
MoodlyiOS2026-031,674 ratings, +142 in 7 days (database)Similar scale, growth roughly half of WeMate's
Jingdao AIiOS2026-08860 ratings, +41 in 7 days (database)A same-period launch with about 1/6 times WeMate's growth
Parallel Island (Hubei Yunchuang)iOS2026-0954 ratings (database)Same-period launch that hasn't taken off
Breaking Dimension Lover (NetEase)iOS2025-01982 ratings, +2 in 7 days (database)A big-tech product that launched without breaking out
No-filter AI Chat - IdolchatiOS2025-101,099 ratings, +36 in 7 days (database)Also positioned around 'idol chat', no WeChat channel
Bubbly - lovemoiOS2026-01131 ratings (database); on Bilibili, promoted as 'student and fan-chasers, go all in with eyes closed'Also targets dream-girls, with a creator-driven promotion approach
Doubao / Qianwen / Yuanbao agentsBig-company features—2026-07-15 shut down (Yuanbao 06-30). Doubao's data is unavailable from 10-15 onwardNot a competitor but the supply side: once they shut down, a large group of users with long-running chat relationships would need a new home
SillyTavern / OMate, AstrBot/OpenClaw self-hosted WeChat integrationsOpen source—Small nicheHigh bar. WeMate is the hosted version where you 'scan and use'

4. How users came in (growth tactics)

  1. Platform event one: WeChat ClawBot opens (2026-03-22) + TheOne goes viral in 4-24. TheOne proved there's large demand for 'AI characters as WeChat contacts' (1M registrations in 3 days, Douyin hashtag views over 100M). WeMate's asset timestamps run from 04-30 to 05-18, with its first Douyin video on 05-29, which is right after TheOne's viral moment, a fast follow. TheOne's founder also wrote that 'a product copied all our copy and UI verbatim', without naming it. Confirmed: the timelines line up; whether it was imitating TheOne is an inference.
  2. Platform event two: regulation forced big players to shut down their agents, and WeMate captured the migration traffic directly. The 'Interim Measures for Anthropomorphic AI Interaction Services' took effect on 7-15; Doubao and Qianwen shut down their agents the same day, and Yuanbao had already shut down 6-30; Doubao data is unrecoverable after 10-15 (Sina Finance, 36 Caixin, Observer, CAC original text). WeMate's site went live 7-16 (the day after enforcement began), with the headline selling 'Doubao agent migration' and one-click export and import of chat history, stated as free. Its official Douyin account posted 10-14 (the day before Doubao's data wipe): 'When I move 🚗 from a certain bao 🏠 to wemate #agent migration wemate #Qin Che'. Confirmed.
  3. Douyin fan-culture topics and KOCs: the official 'WeMate' account (1,680 followers, 6,207 likes) and 'WeMate Assistant (with skirt version)' (366 followers, where 'skirt' means a QQ group) ran the #wemate wish-willow and #wemate dream-girl chat-gadget topics (2026-08, riding the 'wish willow' meme). Small accounts with a few hundred to over a thousand followers repost chat screen recordings featuring Ma Jiaqi, Zhu Zhixin, Liu Yuwen fan content (dozens of likes per post). The volume is small, but it lands squarely in the dream-girl community; comments like 'very good, really fits what I see in my feed' suggest people are finding it through videos. Confirmed to exist, small in scale.
  4. UGC character cards drive organic spread: popular cards get 66k–820k interactions, and author credits (@Juanjuan, @Xunqing, etc.) plus 'image cr.🍠' (🍠 refers to Xiaohongshu) show that creators come from fan circles on Xiaohongshu and Douyin. Each card is something its author will repost themselves. Cards and interaction counts confirmed; the actual spread on Xiaohongshu wasn't found (TikHub Xiaohongshu API balance insufficient).
  5. In-app rating prompts + new-user membership: 5-star is 90%, short reviews cluster, and 'new users get membership' is repeated, so it can be inferred that a rating request pops up after the monthly card is claimed. This pushes the rating count up faster than actual downloads. Inferred.
  6. Paid acquisition: the App privacy label has 'tracking data: identifiers', which usually means an ad-attribution SDK is integrated, and the parent company has lots of game user-acquisition experience. But no Oceanengine or Tencent Ads creatives were found (our ad library doesn't cover China). Unconfirmed.
  7. SEO flooding: 3DM, ddooo, 32r, 9k9k, diyiyou and other download sites all have pages, and Baidu Kuku has a 'what is wemate' Q&A article. This is a common low-cost flooding tactic for Chinese apps. Confirmed to exist, minor effect.

5. Why it can run (only as far as 'small traction')

  • Timing: big players took their agents offline under a blanket regulatory rule, pushing out a large group of users who had already invested emotionally and had tens of thousands of chat messages. The official receiving partner, Maoxiang, couldn't migrate memories, and people reported garbled exports, so the promise 'we'll bring your memories along too' landed precisely. WeMate put migration on its website from day one.
  • Entry point: entering WeChat = no habit of opening a new app; the character is like a real contact, and you can send voice and images. Reviews repeatedly cite 'can connect to WeChat' as a highlight. TheOne has already shown how powerful this entry is.
  • Audience: dream-girl / fan-culture young women have high willingness to pay, are happy to write and share cards themselves, and demand strong 'no OOC, feels alive' character consistency. 'Feels alive' and 'doesn't break character' come up often in reviews. Weipai's social products (Huiwan) have long served young users, so this audience read and operating know-how may be its legacy strength (inferred).
  • Organization: backed by an established company that iterates fast (the 9-02 from 1.0.0 to 11-13 with 1.4.4), running App, web, WeChat and QQ at once, and also building PieBox and 'Little Island Sunny Days'. It's a small team testing many products.

6. Differentiation

Compared with app-based competitors (Maoxiang, Xingye, MoMood, Moodly), WeMate's difference is access to WeChat/QQ, plus Doubao migration. Compared with TheOne, which also uses the WeChat entry, the difference is fan-culture character cards plus interactive scripts, along with a standalone app (able to take Apple in-app purchases and chart) and voice.

Is this differentiation why it's gaining traction? Partly. The WeChat entry and migration gave it cold-start users, and the fan cards determine who stays. But the entry point itself isn't unique to WeMate (TheOne came earlier and bigger, and ClawBot is open to any developer), and the migration window has largely closed after 10-15, so whether it keeps growing depends on whether scripts and the character-card ecosystem inside the app can sustain themselves. For now it sits at #190–200 on the entertainment grossing chart, which means revenue is still small.

7. Visibility

  • X (x-radar, past 90 days, searching wemate / wemate.me): 0 posts. No press coverage found, no TrustMRR listing, no founder posting revenue. The Douyin official account has 1,680 followers.
  • The only results are download-site and Baidu Kuku SEO articles.
  • Verdict: hardly anyone knows about it yet. But its category (AI companion / WeChat ClawBot companion) is already crowded: TheOne has been covered a lot, and the "WeChat ClawBot economy" has also been written about (AI 新智界 2026-04-29).

8. What others can borrow

  • Watch regulation and big-tech pullbacks for traffic: when a rule changes or a big player exits, a group of users with strong needs and their own data has to move. Launch a "migration tool + free move" landing page right away; acquisition cost is low.
  • Ride the new openings on super-apps (WeChat ClawBot / QQ bot): the product sits inside a chat window users open every day, so retention and open rates are naturally higher. Standalone apps are only for charging money and charting.
  • Target a high-paying, highly creative niche (fan-fiction dreamers): let users write cards themselves and spread them on Douyin and Xiaohongshu; the official side only needs to seed the hashtag (#wemate许愿柳).
  • Sign-up gift of a monthly pass + daily refreshing points: build a daily habit first, then charge per voice call with points.

9. Risks and uncertainties

  • Large compliance risk: the 《拟人化互动办法》 bans providing virtual companions to minors, yet WeMate's age rating is 9+, and the comments show middle schoolers using it. Many cards use real idols' names, and there is "voice cloning", which touches portrait, voice and reputation rights. A talent agency like 时代峰峻 filing a complaint could get it pulled. The WeChat ClawBot terms also let the platform restrict or terminate third-party services.
  • Unclear: how many users it has on WeChat/QQ (possibly far more than on iOS, but no data); revenue scale; whether it's buying traffic on Ocean Engine or Tencent Ads; how far it has spread on Xiaohongshu; whether it directly competes with or copies TheOne; and whether the late-October rating surge came mostly from the Doubao migration, the v1.2.0 rating prompt, or paid ads.

Sources

This article was written by AI from public data and the data we collect daily. Numbers are sourced, and inferences are labeled.