Last week, Alibaba launched Qwen3.7-Plus, its latest multimodal agent model, which ranks fifth globally on Vision Arena. The company also unveiled a Qwen-powered swine disease diagnosis assistant with Muyuan Group, while its AI development platform Model Studio released an open-source CLI enabling AI agents to tap into more than 150 multimodal models across text, image, video, and audio.
Alibaba has unveiled Qwen3.7-Plus, its most capable multimodal agent model to date. This new model unifies vision and language into a single, versatile agent foundation. It delivers comprehensive upgrades in vision-language capabilities while retaining robust agentic strength in coding, tool use, and productivity workflows. It is currently ranked fifth globally on Vision Arena, an independent benchmarking platform for evaluating Vision-Language Models (VLMs) based on human preference.
A key feature of Qwen3.7-Plus is its ability to operate as a multimodal interactive hybrid agent, seamlessly blending GUI and CLI interactions within a single agent loop. It functions as a powerful visual agent that can perceive real-world scenes, read screens, operate GUIs, generate codes from visual references, navigate mobile apps end-to-end, and answer visual questions grounded in web knowledge. Beyond static images, the model also enhances video and driving-scene understanding, laying a critical foundation for real-world multimodal agents, autonomous driving and embodied AI scenarios. As a versatile coding agent and productivity assistant, it handles the full spectrum of tasks—from frontend prototyping to complex software engineering and multi-step workflow automation with full-modality input. The model is now available globally for businesses and developers through Alibaba Cloud’s Model Studio, a one-stop model service platform.
Qwen3.7-Plus exhibits excellent performance across benchmarks, showcasing robust coding, tool-use and planning capabilities, with particular strength in complex multi-step planning and GPU kernel optimization.
Alibaba unveiled its latest LLM Qwen3.7-Max last month, which was ranked third globally on ITBench-AA, the pioneering benchmark for agentic enterprise IT tasks, and it achieved fourth place in Code Arena as the top-ranked Chinese lab on the board, on par with Claude Opus 4.6 on agentic web development tasks.
Muyuan Group, a global leader in livestock farming, has partnered with Alibaba Cloud to build an intelligent swine-farming AI model, leveraging Alibaba’s Qwen LLM and advanced computing capabilities. This partnership aims to accelerate AI applications across key operational areas, including feed nutrition, breeding stock improvement, and livestock management.
As a core milestone of this collaboration, an AI agent powered by Qwen was developed to diagnose swine diseases. By synthesizing 18 types of structured data—including pig posture, medical history, and environmental conditions—with the professional expertise of Muyuan’s frontline veterinarians, the agent provides thorough assessments of respiratory, digestive, and infectious diseases. It automatically generates comprehensive reports featuring diagnoses and cost-benefit analyses to support clinical decision-making.
Building on this agent, the partners launched “Xiaomu Assistant,” an AI-powered application that significantly reduced the health inspection time for a batch of approximately 600 pigs from 20 minutes to just seconds. The assistant supports multimodal inputs, enabling farm workers to upload voice notes or take photos of pigsties directly via a mobile app for real-time diagnostic insights.
To maximize operational efficiency, Muyuan Group has fully migrated its business infrastructure to Alibaba Cloud. Moving forward, both parties plan to co-develop a specialized, intelligent farming AI model by combining Muyuan’s deep domain expertise with Alibaba Cloud’s robust AI+Cloud solutions, enabling the intelligent transformation of the livestock industry.
Model Studio, Alibaba’s one-stop enterprise-grade AI development platform, has released an official CLI (command line interface) for AI agents to seamlessly access the platform’s core AI capabilities, covering text, image, video, audio, directly from users’ agentic tools, to improve efficiency in complex workflows. It is open-sourced on Github and works with mainstream agentic tools including Claude Code, OpenCode, Cursor, OpenClaw, Cline, Qoder, and Qwen Code.
Model Studio released an open-source CLI on Github
Model Studio CLI supports out-of-the-box AI capabilities including text chat, image and video generation and editing, speech synthesis and recognition, with access to more than 150 multimodal models on Model Studio, including Alibaba’s Qwen family models, HappyHorse, and Wan, alongside third-party models such as DeepSeek and Kimi. Setup requires only a command in the agent’s terminal and an API key from Model Studio. Once configured, agents can autonomously orchestrate calls across the AI services from the platform.
The tool is designed to boosting productivity in handling complex workflows in the agentic era, especially in a wide range of content creation scenarios such as e-commerce marketing assets, audio podcasts, storybooks and short-form video production, driven by natural prompts.
This article was originally published on Alizila written by Crystal Liu and Karen Zhang.
1,420 posts | 497 followers
FollowAlibaba Cloud Community - June 3, 2026
Alibaba Cloud Community - January 30, 2026
Alibaba Cloud Community - January 30, 2026
Alibaba Cloud Community - January 4, 2026
Alibaba Cloud Community - September 6, 2024
Kalpesh Parmar - May 12, 2026
1,420 posts | 497 followers
Follow
Alibaba Cloud Model Studio
A one-stop generative AI platform to build intelligent applications that understand your business, based on Qwen model series such as Qwen-Max and other popular models
Learn More
Qwen
Full-range, open-source, multimodal, and multi-functional
Learn More
Alibaba Cloud for Generative AI
Accelerate innovation with generative AI to create new business success
Learn More
AI Acceleration Solution
Accelerate AI-driven business and AI model training and inference with Alibaba Cloud GPU technology
Learn MoreMore Posts by Alibaba Cloud Community
