Tongyi Wanxiang
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 6
- Doubao's 13-Egg Test Exposes Image-Generation Counting Challenges
A viral Weibo test showed Doubao repeatedly claiming it had drawn exactly 13 eggs while the images did not match that count. A Guokr analysis cites research testing 15 image-generation models, including Doubao, Kling, Hunyuan and Tongyi Wanxiang. The study found accuracy fell from roughly 60%–80% for requests of one to five objects to about 10%–30% for six to ten, with nearly all models below 10% for 11 to 15. Informal community tests across several AI systems produced mixed results, so the episode does not establish universal failure.
虎嗅 AI · 🔥 3 - Apple Intelligence set for China test release in late October, without Siri AI
A Going Global weekly digest citing Tencent Technology reported that Apple executive chairman Tim Cook confirmed Apple Intelligence and related Alibaba Qwen models have been approved for a test release to Chinese users in late October. The first features include live translation, writing tools and a photo cleanup tool, but Siri AI is not included in this release. Bloomberg's Mark Gurman previously reported that Apple plans a formal rollout in December with iOS 27.3, with first features broadly matching the 2024 overseas launch.
创业邦 科技 · 🔥 0 - Apple Intelligence to launch in China in December via iOS 27.3, reports say
Apple plans to launch Apple Intelligence in China in December through the iOS 27.3 update, initially with AI features that debuted in the U.S. two years earlier, according to media reports cited by the roundup. Under a dual-partner arrangement, Alibaba will integrate AI capabilities into Apple Intelligence for text and image understanding and content generation across iOS, iPadOS, macOS and visionOS in China, with a Sept. 29 license covering Tongyi Wanxiang and Qwen; unnamed sources said Baidu software will be embedded in Siri and Visual Intelligence, and developers found a “Baidu Visual Search” module in iOS 27 Beta 2 firmware. Apple and Baidu have not publicly confirmed the arrangements.
创业邦 科技 · 🔥 0 - Report: Alibaba Tongyi Wanxiang video model lead Zhang Yingya departs
DoNews, citing insiders via Shenma Pengyouquan, reported that Zhang Yingya, who led Alibaba's Tongyi Wanxiang Wan-series video-generation models, has recently left the company; whether he will start a business or join another company is unknown. The report said Zhang joined Alibaba through campus recruitment in 2015, began in visual recognition and image detection, and later moved to image and video generation and diffusion models as a main lead of the Wan video model. Neither Alibaba nor Zhang has publicly confirmed the report.
DoNews · 🔥 0 - Apple Intelligence, Qwen added to China iPhone filings; December launch planned
Multiple China-market iPhones received updated regulatory filings on September 29 carrying three large-model registration numbers: Apple Intelligence (Shanghai-AppleZhiNeng-202506160057), Tongyi Wanxiang (Shanghai-TongYiWanXiang-202410090024) and Tongyi Qianwen (ZheJiang-TongYiQianWen-20230901), according to IT之家 and cnBeta. The filings cover iPhone 15 Pro / Pro Max and later models, plus iPhone Air and the foldable iPhone Duo — matching the device list that supports Apple Intelligence. Apple Technology Development (Shanghai)'s Apple Intelligence model was registered on July 8, 2026 for use on Apple phones, and Alibaba has said Qwen will be integrated into Apple Intelligence to bring text and image understanding and content generation to Chinese users across iOS, iPadOS, macOS and visionOS. Bloomberg reports Apple plans to launch Apple Intelligence in mainland China in December via iOS 27.3, without Siri AI features.
IT之家 AI · 🔥 0 - Meitu's AI turnaround meets cost and competition pressure
An analysis republished by TMTPost from Juchao WAVE says Meitu's shares have fallen from above HK$9 in early 2026 to around HK$4, more than 50% below their peak and down over 40% year to date, as market sentiment toward its AI applications turned cautious. Citing the company's interim results, it reports H1 2026 revenue from continuing operations of RMB 2.213 billion, up 22.1% year on year, and adjusted net profit attributable to shareholders of RMB 652 million, up 39.5%; imaging and design revenue reached RMB 1.769 billion, nearly 80% of total revenue, with more than 18.44 million paid subscribers at end-June, up 19.7% (6.5% penetration). The piece flags rising pressure: editing and image-generation features are becoming free and generic across CapCut, Canva and Google's Nano Banana, H1 2026 compute and cloud costs rose 25% to RMB 134 million (over half of the increase from inference), gross margin slipped 3.8 points to 71.5%, and Meitu is pursuing a 'model container' approach that uses Alibaba's Tongyi Wanxiang base model for video generation.
钛媒体 · 🔥 0
Experience and discussion from the community
Share my Tongyi Wanxiang experienceAsk about Tongyi Wanxiang
Nobody has shared their experience with Tongyi Wanxiang yet.