---
title: "Giants Sell Games to Buy Agents, 3D Modeling Keeps Rising: July's Going-Global AI Traffic Data"
author: "Unique Research"
sourcePublication: "Unique Research Substack"
originalPublishedAt: "2026-08-18T09:13:21+00:00"
canonical: "https://ffcap.cn/en/research/src-20260818-01html"
source: "https://uniqueresearch.substack.com/p/src-20260818-01html"
language: "en"
---

# Giants Sell Games to Buy Agents, 3D Modeling Keeps Rising: July's Going-Global AI Traffic Data

[![](https://substackcdn.com/image/fetch/$s_!ENOf!,w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F01086d6c-ee44-4344-bf44-e124689bb026_2048x1152.png)](https://substackcdn.com/image/fetch/$s_!ENOf!,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F01086d6c-ee44-4344-bf44-e124689bb026_2048x1152.png)

_Original · Unique Research · 2026-08-18_

_Editor's note: The first-person analysis and all data figures belong to the original Chinese author. This English rendition retains the five analysis sections, closing outlook, and methodology notes. All named companies, products, and figures are preserved. Traffic data, deal values, and ARR figures are as reported by Unique Research, not independently verified for this rendition._

AI Industry Observer

Giants Sell Games to Buy Agents, 3D Modeling Keeps Rising

"Going-global AI doesn't lack users anymore — it lacks reasons for users to stay."

The most thought-provoking AI news of these past two days are two seemingly unrelated stories.

On August 17, Alibaba officially announced selling its game business Lingxi Games (灵犀互娱) to Xinyun Capital, with the deal reportedly valued at over $1.5 billion (about 10.1 billion RMB). Lingxi CEO Zhou Bingshu's internal letter was blunt: this is "Alibaba's overall layout based on strategic focus." Translated: sold games, going all in on AI. ByteDance performed the same script three months ago — selling its overseas game core asset Moonton to Saudi Arabia's Savvy for $6 billion. Giants are collectively doing the same thing: turning assets not adjacent to AI into cash.

The other story is Manus. After officially announcing independence on August 11, its first major move wasn't a new product — it was a user letter: partial data generated after December 29, 2025 — the day it was acquired by Meta — will be deleted, and affected users must back up themselves before August 23. A $2 billion "divorce" landed on every user clicking "export data" themselves. Netizen comments were as sharp as ever: "a remnant of the old era."

But the story was told earlier by July's going-global traffic data.

We statistically analyzed the July 2026 going-global AI application chart (100 products, measured as Chinese or Chinese-entrepreneur-led AI Web products targeting overseas markets). The conclusion first: the June panic of "63 products all declining" has stopped, but going-global AI has officially entered stock competition — the bar to stay at the table is higher than last month.

Overall: From 63 Declining to 50/50

Last month's most eye-catching number was "63 declining." This month, the rising/falling split returned to 50 up, 50 down. Total visits grew from June's 492 million to 501 million, +1.9% MoM — the cooling has stopped.

But don't get optimistic too fast; two details show the ceiling hasn't changed:

Top concentration rose, not fell. The top 10 took 47.8% of all traffic (June was 47%), top 5 took 33.3%. What stopped declining wasn "everyone gets a share" — it was the top that stopped, while the mid-tier is still churning.

#1 PolyBuzz showed cracks. Visits 65.96M, +0.87% MoM, basically flat, still #1, still roughly twice #2 Qwen (33.06M). But its MAU dropped 10.1%, from 4.31M to 3.88M — people didn't come less, but fewer stayed. Visits flat, MAU down — this is the most classic signal of a product entering plateau.

"June was panic, July was divergence. The combination of slight total growth and even rising/falling splits translates to one sentence: going-global AI doesn't lack users anymore — it lacks reasons for users to stay."

3D Modeling: Second Consecutive Month All Up, and Accelerating

Last month we saw 3D modeling as "the only sector with all products rising," and specifically discussed whether it was a single-month pulse. July's answer: no.

Meshy 11.13M, +17.1%; Tripo 7.31M, +8.4%; Rodin 6.44M, +111.5% — the three main products all rose again. The sector total 25.53M, +28.2% MoM, faster than June's +15.5%.

Two products to look at individually.

Meshy's growth narrowed to +6% last month, and we judged it "entering plateau." The plateau only lasted one month — July re-accelerated to +17.1%, MAU 5.52M (+13.0%). Globally, only 6 products above 5M MAU still grew double digits in July; Meshy is one of them. After the ~$400M Series B announcement (the largest single deal in AI 3D to date), the traffic curve proves capital didn't misread.

Rodin (Yingmou Technology / 影眸科技) was even more dramatic: +111.5%, doubling for the second consecutive month. Longer view: May ~1.3M, June 3.04M, July 6.44M — nearly 5x in two months. MAU also doubled synchronously (3.39M, +107.9%). After Rodin Gen-2.5 launch, official disclosed subscription user and ARR MoM growth both exceeded 400%, with overseas revenue at 70-80% — Web traffic and commercialization data interlock, and this kind of growth is the most solid.

"Last month we said 3D is moving from content category to infrastructure; July's data turns this from judgment into trend. Games need 3D assets, e-commerce needs product models, robots need simulation data — demand comes from all directions, which is why it can rise across the board and accelerate."

Emotional Companionship: Emochi Cracks Into Top 3, but the Throne Is Wobbling

Emotional companionship remains the #1 going-global category: 7 products totaling ~128M visits, 25.4% of the top 100's total traffic, sector +9% MoM, rising for the second consecutive month.

But the internal seating changes are much more interesting than the totals.

The most aggressive is Emochi (under FlowGPT): visits 25.55M, +31.6%, second consecutive month of big gains (June +58%), directly cracking into #3 overall — only PolyBuzz and Qwen are ahead. Same company, but FlowGPT's main site dropped 20.4%. One company's two curves — one up, one down: FlowGPT is pivoting entirely from "prompt community" to "companionship product," the transformation written in the data.

The mid-tier is also pushing up: HiWaifu +89.4% to 1.78M, Joyland +17.4% (second consecutive rebound), CrushOn +15% to 22.4M, MiniMax's Talkie slightly up 1.8%.

But as mentioned, PolyBuzz's MAU dropped 10.1%. Top stagnating, mid-tier pushing — the sector has no bad news, but the list of who eats everything is quietly changing.

"Companionship in July validated our judgment from last month: no launch events, no grand narratives, but stable demand and high stickiness. The first stable money AI makes going global is still 'emotion.' But the people making this money are shifting from just PolyBuzz to a crowd."

Agents: Manus Down Three Months, the Rising Ones All Have "Backing"

Back to the Manus at the beginning. July continued double decline: visits -15.7% to 19.57M (June was -17.2%), MAU was worse at -22%, from 7.41M to 5.78M. In the seven and a half months after Meta's acquisition, this curve never lifted its head.

That -22% also brought a landmark event: going-global MAU #1 changed hands. Manus gave up the seat it had held for a long time; taking over is image generation's SeaArt AI (MAU 6.1M, +9%). An Agent product's MAU throne was taken by a "generation" product.

So August 11's "restoration of independence" has a clear meaning in data terms: this isn't a triumphant return, it's damage control. $2 billion bought the qualification to re-enter the board, but that sharp question didn't come with the shares — after model vendors go directly into Agents, how much value is left in the middle layer?

But there's a contrast worth pondering; fairness requires noting: according to media reports, during the half year in Meta, Manus' annualized revenue run rate grew from ~$100M to $400-500M. Web traffic declined for seven and a half months, but revenue quadrupled — it lost the onlookers and kept the paying users. So "damage control" is only half right: Manus' Web traffic story is indeed over, but the subscription revenue business actually got off the ground. The question after restoring independence is therefore more specific: can it pull the traffic curve back too?

The control group in the same track is Z.ai (Zhipu / 智谱): visits 19.89M, +20.4%, second consecutive month of big gains (June +36.9%) — this month it edged Manus out by 320K. Going-global visits #5, agent track #1 — the throne has changed hands.

Another one surging: MuleRun, visits +147.9% to 1.67M, MAU +138.1%. But its growth isn't from product breakout, it's from "consolidation" — on August 3, it and QoderWork and Wukong were integrated by Alibaba into "Qwen Office" for public beta, led by new DingTalk CEO Chen Yusen. MuleRun's curve is essentially a signal of Alibaba pulling its free-range Agents back into the main fleet. Combined with the opening news: August 17 selling Lingxi Games for 10B RMB cash. Selling games with one hand, absorbing Agents with the other — where Alibaba's重心 is moving is written clearly.

"The July agent track pattern can be said in one sentence: the rising ones all have backing. Z.ai is backed by Zhipu's model, MuleRun by Alibaba's office scenarios, Manus used to have Meta. The independent 'general Agent' narrative is receding; the 'model vendor + specific scenario' combination is entering."

ByteDance's July: Video, Code, and Homework Help All Down

If you rearrange the July chart by corporate group, the most pressured company is ByteDance.

Dreamina, image generation, visits -17.9% to 10.1M, MAU -20.3% — last month it was the +23.5% star student, this month it reversed directly. TRAE, coding tool, -16.2%, second consecutive month of big decline (June -25.5%). Gauth, photo homework search, -31.9% to 2.72M, MAU -30.9% — of course there's summer seasonality, the real verdict is in September's back-to-school season. The only decent one is Dola, visits +6.9% holding 10M, but MAU also slightly down 3%.

Each product has an explanation individually, but together it's no coincidence. March selling Moonton to Saudi Savvy for $6B, end of July splitting the Feishu team in two into Doubao and Volcano Engine — the organization is consolidating, traffic is bleeding, and the two lines tell the same story: ByteDance's AI products have ended the laying-out phase, but the flagship product hasn't caught the receding share yet.

Growth Chart Deep Read: One Real Case, One Small Base

July's going-global growth chart head — we checked backgrounds one by one, conclusion: no caliber accidents, but quality needs to be separated.

The real growth representative is ChatCut (+293.5%, 360K). Small volume, but the product implication is the richest this month: it's a browser-based AI video editor. In July it did two things — released official plugins for OpenAI Codex and Claude Code, exposing the entire editing, subtitle, effects, voiceover, export workflow as an MCP tool set. The effect: you type a sentence in Codex, and the Agent directly cuts on ChatCut's timeline. This gameplay went viral in the developer community in mid-July, and that's where the traffic came from.

Why single it out? Last month we judged "video editing is dying," and ChatCut gave that judgment a more precise version: what's dying isn't video editing, it's video editing that Agents can't call. Same tools — those that transform themselves into Agent hands can jump from the being-eaten list to the growing list.

Growth chart #1 Imbue (+393.7%) needs a cool head. Absolute value only 150K, small-base amplification effect. This is a San Francisco AI lab (founder Kanjun Qiu is Chinese-American, former Dropbox exec), doing coding Agent product Sculptor — not a traffic product, so mark this increase as "pending verification" for now.

Further down, Bloome (+87%, 140K), Elser AI (+57.3%, 230K), Looki L1 (+47.4%, 260K) are all sub-300K small bases, also just noted without expansion.

Self-Calibration: Two Judgments From Last Month Need Correction

A data organization's credibility is built by raising its own hand when it's wrong. Two sectors in July moved differently from our June judgment.

Video editing didn't collapse. June this sector totaled -11.8%, Wondershare's three products all down, and we judged "independent post-production tools are being eaten by the generation stage." July: Media.io went from -22% to +2.7%, OpusClip +5.2%, Meitu's Vmake +10.9%. The stabilization reason, combined with the ChatCut case, might be that these tools are also rushing to attach Agent capabilities. Judgment revised: video editing elimination is delayed, but the rules haven't changed.

Smart search stopped declining. June we said "independent AI search is being nestedly absorbed," when Genspark fell 43% in two months. July: Genspark +1.1% stopped the decline, Alibaba's Accio +2.4%, felo visits +18.9%, MAU +31.1%, flowith double metrics positive. One month isn't enough to overturn, but enough to downgrade the "second disruption" conclusion to "pending observation" — search demand definitely still exists; whether these products have found their ecological niche, give it two more months.

Image editing completely followed last month's script: Fotor -10.3%, Airbrush -17.3%, insMind -14%, continuing to decline. Mind mapping also still down (Xmind -6.3%, Mapify -16.8%). Music duo Mureka -6.4%, Udio -12.5%, copyright shadows not dissipating — this industry-wide trouble continues.

Three Hanging Questions From Last Month All Got Answers in July

At the end of the June article, we left three things. July's data gave all the answers:

opencode settled in. 11.59M, +6.7%, MAU 3.42M (+4.1%), overall chart #10. The open-source coding tool position — it went from "dark horse" to "resident."

Agnes' free storm receded. June rushed to +570% via free multimodal API; July visits -22.6%, MAU -20.5%. Free can buy one month of screams, not the second month's retention — the "price butcher's" bill arrived in the second month.

Education AI still at the bottom. Gauth declined two consecutive months, but seasonality is the main factor; September's back-to-school season is still the final judge.

August: Watch Four Cards

First card: Manus' first full month after independence. August's chart will answer whether "restoration of independence" is stopping bleeding or last gasp. By visits it's already been overtaken by Z.ai; by MAU it lost the seat too — the first thing after re-entering is to take back both thrones.

Second card: Qwen Office public beta. MuleRun +147.9% is just the appetizer; the four-in-one Qwen Office backed by DingTalk's enterprise clients is Alibaba's first battle to "recover C-end lost share from B-end."

Third card: September back-to-school countdown. Gauth's return-to-school curve decides whether education AI is cyclical fluctuation or structural decline.

Fourth card: Can 3D modeling keep accelerating? Sector up all members for two consecutive months, MoM from +15.5% to +28.2% — if August continues, this is the first sector-level trend in 2026 going-global AI that's been repeatedly validated.

The overall numbers already speak clearly: even rising/falling, slight total growth, tighter top. The second half of going-global AI — the window for telling stories is over. Now it's two things: either be close enough to real production processes like 3D modeling, or like emotional companionship, make a specific emotion indispensable.

August chart, see you in a month.

Unique Research Rankings Release

Unique Research — Authoritative AI Market Insights Institution

Unique Research's July 2026 AI going-global Web ranking details are as follows:

AI Going-Global Website Traffic Chart: TOP 100 by visits and MAU for going-global AI Web products.

All ranking charts also provide growth charts, helping investment institutions find promising projects and entrepreneurs and developers find potential opportunities.

Our influence and recognition:

Unique Research's data and insights have been widely trusted and cited by top domestic and international investment institutions, academic entities, and media, becoming an important benchmark for market analysis, investment decisions, and trend judgment. Our partners and citers include (but are not limited to):

Securities firms: Sinolink Securities, China Securities, Huayuan Securities, Haitong Securities, etc.

Authoritative media: China Securities Journal, TMTPost, Jiemian News, etc.

We firmly believe that only data that withstands verification by industry core participants has lasting value.

We believe in the power of open source and openness. Unique Research hereby makes three commitments:

This AI product ranking is permanently open-source and free, with complete methodology and raw data supporting full reproduction.

This AI product ranking's data sources are strictly limited to independent third-party monitoring platforms; self-reported enterprise data is not accepted as evaluation basis.

This AI product ranking's ranking system maintains absolute neutrality, accepting no commercial cooperation or paid practices that interfere with ranking results.

How to use this ranking?

If you are an investor/analyst: this ranking provides you with a verifiable, clean data benchmark — a reliable foundation for building investment models, finding underwater projects, and cross-validation.

If you are an AI entrepreneur/product manager: you can not only precisely locate competitors, but also deeply reproduce our methodology to analyze your own product's market performance and optimize iteration.

If you are a researcher/student: the fully open-source nature makes this report an excellent free resource for academic research and market analysis. We encourage innovative research based on this data.

For complete data behind the charts and rankings shown in this article, please visit 100aiapps.cn. All visible data is copyable and open for free non-commercial use in personal study and research. Commercial use requires prior authorization consultation with Unique Research. We welcome listed teams to apply rankings to business plans, recruitment materials, etc., under compliant use conditions, providing authoritative evidence of your strength.

Join the AI Think Tank Open Source Community

Unique Research not only focuses on AI industry research, but also hopes to build an open-source community promoting AI market data transparency. We invite developers, data scientists, and industry analysts:

Reproduce and verify: visit our Feishu knowledge base for complete data and methodology, verify our results and suggest improvements.

Build together: contribute code to help us expand data dimensions (such as API call volume estimation, agent monitoring models, etc.).

Exchange insights: join our Feishu group to share unique insights based on Unique data.

One person's insight is limited; a group's wisdom is infinite. Let's define the measurement standards for the AI market together.

01\. Going-Global AI WEB Rankings

1.1 Going-Global AI WEB TOP 100 by Monthly Visits

Based on Unique Research's independent AI industry research and continuous tracking of thousands of AI companies and nearly 10,000 AI products globally, sorted by that month's cumulative visits for each AI WEB product serving overseas markets, organized into the top 100 going-global AI WEB products by visits.

1.2 Going-Global AI WEB TOP 100 by Monthly Active Users

Based on the same research and tracking, sorted by that month's active user count for going-global AI WEB products, organized into the top 100 going-global AI WEB products by active users.

02\. Research Statement

2.1 Data Description

The AI product revenue data in this report is based on real-time tracking of redirect traffic from AI product official websites to payment gateways (such as Stripe), combined with time-series analysis, deeply analyzing historical visits, monthly active user fluctuations, and pricing strategy changes to build a dynamic revenue evaluation model. Core model parameters are determined through a triple verification system: paid conversion funnel monitoring, enterprise-disclosed data reverse calibration, and industry benchmark conversion rate comparison, and timely cross-referenced and corrected with enterprise earnings, financing announcements, and authoritative media disclosures. Through sampling verification, the mean absolute error rate between predicted and actual data stably stays within ±10%, reaching industry research-grade accuracy.

The AI product traffic data in this report is based on multi-source data integration and Unique Research's intelligent algorithm processing, with monitoring scope strictly limited to users' direct access through official apps and websites. Specifically, this data only covers users' direct access to AI products through websites and native apps, and does not include traffic from: browser plugins/extensions, desktop client software, WeChat/Alipay mini-program ecosystems, Discord and other third-party embedded services, open-source model local deployments, API calls, and other non-direct access scenarios. This data focuses on AI product core-side access monitoring, aiming to objectively reflect mainstream user terminal direct usage.

Given the complexity and dynamism of data collection and processing, and continuously changing market conditions, the data shown may contain some degree of error and omission. Therefore, this data should be considered a research and analysis reference, aimed at providing users with a window into understanding the AI product market, not a deterministic basis for specific investment strategies or advisory recommendations.

Data coverage expansion note: We are actively developing and testing monitoring models for the following emerging areas, expected to be gradually included in future quarterly reports:

AI Agent ecosystem

AI hardware device sales and activity

Large model API call scale estimation

Stay tuned.

2.2 Concept Definitions

Geographic dimension

Overseas AI products: Products founded by non-Chinese entrepreneurs or teams and primarily launched for the global market.

Domestic AI products: Products founded by Chinese entrepreneurs or teams, primarily launched for the Chinese domestic market.

Going-global AI products: Products founded by Chinese entrepreneurs or teams but targeting overseas markets.

Functional dimension

AI-native applications: Applications that deeply integrate AI technology from the product design stage, whose core value, business processes, or user experience are entirely dependent on AI technology.

AI-feature applications: Applications that enhance existing features by integrating AI technology on top of existing business logic.

AI ecosystem applications: Platform-type applications that provide support, connection, or communication facilitation for the AI ecosystem.

2.3 Metric Definitions

WEB data metrics

Visits: The total number of all visits to a website within a specific time period.

Unique Visitors: The number of independent IPs visiting the target website within a specific time period.

ARR: Within a certain period, the total revenue earned through subscription services, excluding advertising revenue, transaction commissions, and professional services.

APP data metrics

Downloads: The number of times an App is downloaded by users within a specific time period.

Active Users: The number of independent users who performed at least one activity within a specific time period.

IAP: Revenue generated by users purchasing virtual goods, additional services, or premium features within an App, excluding advertising revenue, direct user payments, and third-party Android app store revenue.

2.4 Disclaimer

This report is published by Unique Research, and its copyright belongs to Unique Research. Any Chinese republication or citation must indicate the report source. Foreign institutions requiring republication or citation should contact for authorization in advance. When citing data, please label: \[Data source: Unique Research\]. When directly citing charts, please label: \[Data source: Unique Research, Chart production: Unique Research\].

This report is Unique Research's independent original analysis as a third-party institution. The content does not represent any enterprise's position and does not constitute investment advice to anyone. Investors should note that any investment decisions made based on this are unrelated to Unique Research and its employees or affiliated institutions.

To the extent permitted by law, Unique Research and its affiliated institutions may hold equity in companies mentioned in the report, or provide or seek to provide fundraising or financial advisory services, and its employees may serve as directors of companies mentioned in the report.

---

Original publication: https://uniqueresearch.substack.com/p/src-20260818-01html
On-site reading page: https://ffcap.cn/en/research/src-20260818-01html