{"slug": "alibaba-says-qwen3-8-is-the-worlds-no-2-ai-model-it-has-shown-no-proof", "title": "Alibaba says Qwen3.8 is the world’s No.2 AI model. It has shown no proof.", "summary": "Alibaba previewed Qwen3.8, a 2.4-trillion-parameter multimodal AI model, at the World Artificial Intelligence Conference in Shanghai, claiming it is second only to Anthropic's Fable 5, but provided no benchmark scores, model card, or independent tests to support the claim. The launch follows Moonshot's release of Kimi K3, which rattled US chip stocks, and signals an escalating open-weight arms race among Chinese AI labs. Alibaba shares rose as much as 5.4% on Monday.", "body_md": "Days after a rival Chinese lab shook Silicon Valley, Alibaba has fired back. The company previewed Qwen3.8, its most powerful model yet, and made a bold claim: it trails only one model on Earth.\n\nAlibaba unveiled the model at the World Artificial Intelligence Conference in Shanghai. Its Qwen team said on [X](https://x.com/Alibaba_Qwen/status/2078759124914098291) that Qwen3.8 is “second only to Fable 5,” a nod to Anthropic’s latest release. Alibaba shares rose as much as 5.4% on Monday.\n\n## What Qwen3.8 is\n\nThe model holds 2.4 trillion parameters. That makes it the first in the [Qwen](https://thenextweb.com/news/alibaba-qwen-apple-intelligence-china-approved) family above a trillion to be multimodal, meaning it handles images, video and documents as well as text.\n\nAlibaba says Qwen3.8 should beat its own previous flagship, especially at coding and office work such as full-stack development and data analysis. Lead developer Shuai Bai called it the team’s first trillion-parameter multimodal model, with understanding that he said matches or beats leading proprietary systems.\n\nThe preview is available now. It runs through Alibaba’s Token Plan subscription and its Qoder developer tools, priced at 10% of the standard rate during the trial. Full open weights are promised “soon,” with no date and no licence terms yet.\n\n## A claim without the receipts\n\nHere is the catch. Alibaba offered no benchmark scores, no model card, and no independent test to back the “second only to Fable 5” billing, [SiliconANGLE reported](https://siliconangle.com/2026/07/19/alibaba-previews-qwen3-8-claims-second-claude-fable-5/).\n\nThat is a shift for the company. Its last flagship, Qwen3.7-Max, shipped in May with a full set of published results. This one arrived with none.\n\nThe gap matters because the comparison is Alibaba’s own. No public leaderboard has scored Qwen3.8. On LMArena, where Fable 5 sits at number one, the older Qwen3.7-Max ranks well down the list. The weights, once released, would let outside researchers check the claim. For now, they cannot.\n\n## The open-weight arms race\n\nThe timing is no accident. The launch came three days after Moonshot released [Kimi K3](https://thenextweb.com/news/moonshot-kimi-k3-largest-open-model), a 2.8-trillion-parameter open model that topped a major coding leaderboard and [rattled US chip stocks](https://thenextweb.com/news/kimi-k3-china-ai-tech-rout-selloff). Alibaba’s move reads as a direct answer, [The Decoder noted](https://the-decoder.com/alibabas-qwen-takes-on-kimi-k3-with-open-weight-qwen-3-8-says-model-is-second-only-to-fable-5/).\n\nA cluster of Chinese labs is now shipping models at a scale once reserved for the biggest US labs. Moonshot has Kimi K3. Zhipu has [GLM-5.2](https://thenextweb.com/news/zhipu-stock-surge-china-ai-anthropic-curbs-glm-5-2). All are chasing the frontier, and most are handing out their weights for free.\n\nThere is a strategic wrinkle. Alibaba, like Moonshot, is putting the model in paid products first and releasing the weights later. It turns the launch buzz into customers before anyone can run the model for nothing.\n\nThe bigger signal is what Alibaba is promising to open up at all. Its largest models have not usually been open-weight, a point several analysts flagged. “Something is changing,” the AI researcher Nathan Lambert wrote, reading the shift as a new phase of competition among Chinese labs.\n\nHanding over a claimed number-two system, if the weights match the hype, would push the [open-weight](https://thenextweb.com/news/thinking-machines-inkling-open-weight-model) field deeper into frontier territory. That is the ground US labs have guarded most closely. The proof, though, lands only when the download does. Until then, Alibaba is asking the market to take its word for it.\n\n## Get the TNW newsletter\n\nGet the most important tech news in your inbox each week.", "url": "https://wpnews.pro/news/alibaba-says-qwen3-8-is-the-worlds-no-2-ai-model-it-has-shown-no-proof", "canonical_source": "https://thenextweb.com/news/alibaba-qwen38-second-only-fable5-open-weight", "published_at": "2026-07-20 11:15:42+00:00", "updated_at": "2026-07-20 13:26:54.289718+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-products", "ai-startups", "ai-policy"], "entities": ["Alibaba", "Qwen3.8", "Anthropic", "Fable 5", "Moonshot", "Kimi K3", "Zhipu", "GLM-5.2"], "alternates": {"html": "https://wpnews.pro/news/alibaba-says-qwen3-8-is-the-worlds-no-2-ai-model-it-has-shown-no-proof", "markdown": "https://wpnews.pro/news/alibaba-says-qwen3-8-is-the-worlds-no-2-ai-model-it-has-shown-no-proof.md", "text": "https://wpnews.pro/news/alibaba-says-qwen3-8-is-the-worlds-no-2-ai-model-it-has-shown-no-proof.txt", "jsonld": "https://wpnews.pro/news/alibaba-says-qwen3-8-is-the-worlds-no-2-ai-model-it-has-shown-no-proof.jsonld"}}