Alibaba unveils Qwen 3.7 Max at inaugural Singapore conference
Chinese tech giant debuts new AI model capable of extended autonomous tasks, alongside a major upskilling initiative backed by the Singapore government to ensure no jobless growth in the age of AI
Artificial intelligence (AI) is progressing from a tool that simply summarises knowledge to a digital workforce capable of taking independent action. That was the key message delivered by Alibaba Cloud today as the company hosted its inaugural international Qwen Conference in Singapore, unveiling its latest flagship model, Qwen 3.7-Max, and a suite of agentic cloud products.
Choosing Singapore – the headquarters for its international business unit – as its launchpad, Alibaba Cloud is doubling down on agentic AI, where models act as autonomous agents that can reason, code, and execute complex, multi-step tasks.
“The future of human society will be a human-agent match, meaning humans and agents will work in harmony,” said Li Fei Fei, chief technology officer and president of international business at Alibaba Cloud. “A workforce of agents has arrived, and what that means is an orders-of-magnitude more productive, indefinitely more available workforce at the fingertips of human civilisation.”
To ease the transition to agentic AI, the company unveiled Qwen 3.7-Max, a proprietary model boasting over one trillion parameters and a one-million-token context window. According to Ken Xu, solution architect director at Alibaba Cloud Singapore, the new model is tailored for agentic workflows.
“It can run by itself for 35 hours autonomously without any performance degradation,” Xu noted during a press briefing, adding that Qwen 3.7-Max ranks among the top models on the Artificial Analysis Intelligence Index, closing the gap between Eastern and Western models.
Alibaba Cloud also announced the launch of Qwen Cloud, an AI-native platform designed to serve both human users and AI agents, along with a new skills portal that allows AI agents to convert common cloud capabilities across more than 60 cloud products into skill-based and MCP [model context protocol]-compatible formats.
For enterprise deployments, the company debuted the JVS Agent Suite, providing developers with secure toolkits to build and run AI agents, including OpenClaw agents.
AI that works for workers
However, the use of AI agents has sparked anxieties about workforce displacement. Addressing the conference, Desmond Tan, Singapore’s senior minister of state in the prime minister’s office and deputy secretary-general of the National Trades Union Congress (NTUC), described a distinctly Singaporean approach to the AI revolution: prioritising human capital alongside technological advancement.
“We want to pursue AI, but we also want to pursue AI with no jobless growth – not AI instead of workers, but AI that works for workers,” Tan said. “Our priority is simple: workers must move ahead together with AI.”
A workforce of agents has arrived, and what that means is an orders-of-magnitude more productive, indefinitely more available workforce at the fingertips of human civilisation
Li Fei Fei, Alibaba Cloud
To that end, Alibaba Cloud, the NTUC’s Tech Talent Assembly, and ST Telemedia Global Data Centres announced a joint initiative to equip over 1,000 local small and medium-sized enterprises (SMEs) and students with practical skills in generative and agentic AI.
Starting in June 2026, eligible employees from NTUC union companies will receive tokens to access Alibaba Cloud’s advanced AI tools, including Qwen and its video generation model Wan. Participants can also opt for a subscription to Qoder, Alibaba’s new agentic AI-powered coding platform, alongside access to hands-on training workshops.
“This is what tripartism can look like in the AI world,” Tan added, noting that trust between the government, labour unions, and tech platforms is essential for navigating economic disruption. “Trust makes everything simple, but building trust is anything but simple.”
Tech for good
Besides the product and training announcements, Alibaba Cloud also showcased how Qwen is being used by businesses ranging from enterprises to local startups.
Andy Lee, Alibaba Cloud’s managing director for Singapore and Thailand, pointed out the model’s enterprise adoption across media, gaming, retail and e-commerce, sectors where “we have a big presence”.
To illustrate Alibaba’s regional impact, Lee singled out the company’s collaboration with AI Singapore (AISG) on the Sea-Lion, Southeast Asia’s first large language model. Fine-tuned from Qwen’s architecture, Sea-Lion is being deployed by enterprises across ASEAN to handle complex translations and capture the region’s unique linguistic and cultural nuances.
But Lee noted that the true promise of AI lies in its localised, real-world applications – what he referred to as “tech for good”.
Among the highlighted customers was FingerDance, a startup that translates spoken language into sign language using digital avatars. Powered by Alibaba Cloud’s speech recognition and Qwen models, the platform translates natural language, including Singlish, into real-time sign language annotations.
“For the hard-of-hearing and deaf community, text-based information is not fully accessible because their first language is a visual-based language,” said Kevin Kung, co-founder of FingerDance, noting that the company’s platform aims to bridge the accessibility gap for roughly 500,000 individuals with hearing loss in Singapore.
Other showcased use cases included ArchAIve, a platform that leverages Qwen’s visual capabilities to digitise and preserve faded handwritten Chinese documents and calligraphy dating back to the 1900s, and FizzDragon, an AI-generated content platform that helps creators convert natural language stories into short dramas and video productions using AI agents.
While fully autonomous hacking bots remain a distant reality, an ESET expert warns that AI is quietly supercharging phishing schemes and creating new vulnerabilities inside organisations.
Nutanix CEO talks up the company’s agentic AI play, the growing demand for sovereign cloud capabilities, and why decoupling storage from HCI is hastening migrations from VMware.
Read more on Artificial intelligence, automation and robotics
