Alibaba’s New AI Aims to Rival Meta’s Offerings

Alibaba is intensifying the AI model competition with its new Qwen3.8-27B model, designed for consumer hardware and boasting strong capabilities. By open-sourcing its most powerful model, Alibaba asserts its leadership in the open-weight AI sector, challenging Meta’s initiatives. This move, coupled with on-device AI advancements, positions Alibaba as a significant global competitor to U.S. tech giants.

Alibaba's New AI Aims to Rival Meta's Offerings

A man walks past the Alibaba logo displayed at its booth during the China International Supply Chain Expo in Beijing, China July 16, 2025.

Florence Lo | Reuters

Meta Platforms is making a significant play to lead the open-source artificial intelligence landscape, a domain where Chinese tech giant Alibaba has just served a potent reminder of the intense competition. On Monday, Alibaba unveiled a new AI model engineered to operate on consumer hardware like laptops, signaling a strategic shift and intensifying the battleground for AI model developers. Simultaneously, the company opened the weights of its most powerful model, underscoring its commitment to maintaining a dominant position in open-source technology.

This move by Alibaba follows closely on the heels of Meta’s recent announcement to open-source its most advanced AI model and introduce new variants designed for laptop deployment. These initiatives are part of Meta’s broader strategy to position itself as a compelling U.S.-based alternative to Chinese technological advancements, aiming to differentiate from AI research labs like OpenAI and Anthropic, which favor a more proprietary development approach.

Alibaba’s newly launched model, Qwen3.8-27B, has been lauded for its “excellent capabilities” in code generation, professional tasks, research endeavors, and complex agentic operations. The company claims it rivals the performance of models ten times its size. Furthermore, Alibaba has made the weights of its flagship Qwen3.8 Max model freely available. These weights, which encompass the intricate calculations and rules governing the AI’s functionality and behavior, can now be downloaded and run by anyone. While the model’s architecture and training methodologies remain undisclosed, the open-weight release itself is a significant step.

Alibaba has carved out a leadership position in the open-weight AI sector, with other Chinese companies such as DeepSeek and Moonshot also emerging as formidable contenders.

Meta is not well positioned in the AI race, says Elevation Partners' Roger McNamee

Meta, an early pioneer in the open-source AI space with its Llama series of models, has found itself outpaced by Chinese competitors. Its recent announcement included the Muse Glimmer family of models, also open-sourced and specifically designed for laptop use.

“Meta’s own re-embrace of open weights… was itself a response to two years of Chinese labs… taking a large share” of the open-weight market, Nick Patience, AI lead at the Futurum Group, told CNBC.

Alibaba’s Strategic Push for Dominance in Open-Source AI

The traction and influence of open-weight models are largely gauged by their download counts and the extent to which developers integrate them into their own product development. Hugging Face, a leading repository for open-weight AI models, reported last week that Qwen-based models have generated a staggering 151,448 derivative instances – signifying downloads and subsequent usage by developers. This figure represents 2.6 times the total footprint of Meta’s offerings on the platform, according to Hugging Face data.

Alibaba's Qwen surpasses rivals in downloads, 'tipping point' for investors rotating in China AI

“The company which can offer the most capable open weights models will move ahead in this race,” Neil Shah, co-founder at Counterpoint Research, told CNBC. “Alibaba aims to become this undisputed leader, outpacing Meta and eyeing the global market… as a strong alternative to Silicon Valley frontier-grade deployable models.”

The introduction of Qwen3.8-27B strongly suggests that Alibaba anticipates a future where more sophisticated AI models will operate at the “edge” – directly on user devices – rather than solely within centralized data centers. Industry experts highlight that on-device AI offers distinct advantages, including potentially faster processing speeds and enhanced security due to local data handling.

Counterpoint Research’s Shah identifies the ability of AI models to effectively run on-device, as opposed to relying on cloud infrastructure, as the “next battleground.” Futurum Group’s Patience further elaborates that Alibaba has strategically built advantages across multiple fronts, encompassing open-weight models and the burgeoning field of on-device AI.

“Alibaba has made Qwen the most credible non-US model family to build hardware relationships around, in China and in the open-weight developer community globally,” Patience added, emphasizing Alibaba’s growing influence and ecosystem development.

Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.

Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/24939.html

Like (0)
Previous 1 day ago
Next 2026年4月30日 am1:26

Related News