appConfigEntityId

bodyType

customContents

entityId

entityUuid

isVideo

articlePageQuery

article

articleSchemaContent

content

hooksArticlePageAdvertisementConfigArticle

hooksArticlePageFloatingButtonConfigArticle

appConfig

filter

json

liveArticle

Xinmei Shen

Anthropic’s powerful new Claude Fable 5.1 model has widened its lead in performance benchmarks over Chinese rivals, even as budget-friendly open-weight models from China continue to gain commercial traction globally.
The American company said on Wednesday that Fable 5.1, along with a restricted-access version called Mythos 5.1, was the world’s most advanced model for software coding and complex knowledge work.
The standard version scored 66 on Artificial Analysis’ Intelligence Index, taking the top spot globally and pulling six points clear of Moonshot AI’s Kimi K3 and Z.ai’s GLM-5.3, both leading Chinese models.
Fable 5.1 also claimed first place on San Francisco-based Vals AI’s index for handling complex, real-world tasks across sectors such as finance, coding and law. Anthropic’s earlier Opus 5 and Fable 5 models trailed closely in second and third place.
The benchmark results prompted industry insiders to weigh in on whether leading labs in the United States were pulling further ahead of Chinese competitors.
Yuchen Jin, a technical staff member at US artificial intelligence platform Databricks, called Fable 5.1’s capability leap “insane”, while some observers noted its performance shattered claims that Chinese open-weight developers had already closed the gap with the frontier.
However, a stark price gap points to a sharp divide between the two countries’ approaches.
Despite Anthropic slashing cache read prices by 75 per cent, reducing overall user expenses by up to 45 per cent, Fable 5.1 still costs US$3.69 per task on the Intelligence Index, compared with just US$0.84 for Kimi K3 and US$0.68 for GLM-5.3.


Cached reads refer to the process in which an AI model reads inputs that have already been processed and stored.
“That’s where I think the market starts to split naturally,” wrote Tory Green, co-founder of AI infrastructure firm io.net, on social media on Wednesday. “Cheap open models eat more of the routine agent work, while Fable gets called when the task is hard enough to justify the premium.”
Chinese AI developers, meanwhile, continue to scale up performance.
Tech giant Alibaba Group Holding, owner of the South China Morning Post, released Qwen3.8-Max-0902 on Wednesday. The upgraded model topped Arena AI’s benchmark for front-end web development tasks, beating Opus 5 and Kimi K3.
Meanwhile, Z.ai founder Tang Jie told investors this week that its coming GLM-6.0 model aimed to achieve recursive self-improvement – a capability that would allow it to autonomously train its own successor. Vals AI CEO Rayan Krishnan said on social media on Wednesday that Fable 5.1 had already shown strong signs of this capability in early testing.
As the US-China AI race accelerates, safety and oversight concerns are also intensifying.
Anthropic rival OpenAI teased its coming Astra model on Wednesday, stating that it featured “a critical level of cybersecurity capability” that would enable the system to discover and exploit unknown vulnerabilities across well-protected systems without human intervention.
While OpenAI promised strict guardrails and restricted access, safety researchers expressed alarm after The Information reported on Tuesday that Astra used a “recurrent depth” architecture – a design that boosts performance but obscures how a model reaches its conclusions.
Ryan Greenblatt, chief scientist at Redwood Research, warned the technique impaired oversight, saying it “may be the single worst development for AI security [and] safety to date.”


Philippine tribe vows to stand ground against US-led Pax Silica: ‘listen to us’

As the US weighs restrictions on cloud computing, how will China’s AI sector adapt?

Chinese AI chipmaker Hygon plots expansion from data centres to robotics

For China’s hi-tech IPOs, it’s Hong Kong vs the mainland – but how much does it matter?

Artificial intelligence (AI)

Frontier AI at a cost: what Anthropic’s Fable 5.1 means for the US-China model race

Anthropic, based in San Francisco, specialises in developing advanced AI models. Photo: Reuters

Call for Trump-Xi AI safety talks comes as state media slams US AI governance

The OpenAI logo is seen in this illustration created on June 11, 2026. Photo: Reuters

The Anthropic logo is seen in this illustration taken May 20, 2024. Photo: Reuters

Congress advances new bill to give US edge in open-source AI race with China

An influx of powerful new open-weight models, including Moonshot AI’s Kimi K3, has helped China dominate the industry recently. Photo: Reuters

Nvidia CEO Jensen Huang speaks during the AI Ecosystem event in Tokyo, Japan, on July 16. Photo: Reuters

Sony Music, Warner accuse Anthropic of ‘blatant theft’ in major new lawsuit

Anthropic is facing a fresh lawsuit over its use of music to train its AI models. Photo: Reuters

MiniMax revenue surges 283% but remains behind pace to meet forecast amid crowded AI race

MiniMax is known for its M-series large language models and popular consumer apps. Photo: Shutterstock

DeepSeek leads surge in low-cost Chinese open-weight models on US platform

The DeepSeek app is seen on a smartphone screen in this arranged photo taken January 26, 2025. Photo: Shutterstock Images

People visit the Moonshot AI booth, featuring the Kimi K3 model, during the World Artificial Intelligence Conference in Shanghai. Photo: AFP

Tech

Big Tech

Tech War

Tech Trends

News

China Future Tech

The US firm’s new models claim top position on leading benchmarks, though steep operating costs against Chinese rivals highlight a growing divide


The US firm’s new models claim top position on leading benchmarks, though steep operating costs against Chinese rivals highlight a growing divide.


Frontier AI at a cost: what Anthropic’s Fable 5.1 means for the US-China model race

.css-1c6uqr6{color:inherit;font-weight:inherit;font-size:inherit;font-family:inherit;line-height:inherit;overflow-wrap:break-word;}The US firm’s new models claim top position on leading benchmarks, though steep operating costs against Chinese rivals highlight a growing divide

The US firm’s new models claim top position on leading benchmarks, though steep operating costs against Chinese rivals highlight a growing divide