Alibaba’s Qwen Team Launches Qwen3.7-Plus, Adding Vision, Deep Reasoning, Tool Invocation, and Autonomous Iteration on the Bailian Platform
Alibaba's Qwen team has launched Qwen3.7-Plus, a new multimodal large language model available on the Bailian platform. This advanced model excels in understanding images and video, alongside written prompts, and integrates advanced capabilities like deep reasoning and autonomous iteration to perform complex tasks. It represents a significant step towards multimodal hybrid agent technology, allowing the model to plan and act across multiple steps, making it a robust tool for various applications.
Alibaba's Qwen team has released Qwen3.7-Plus, a new multimodal large language model now accessible via Alibaba Cloud's Bailian platform. This launch follows the unveiling of the Qwen3.7 generation in May and provides API services to external developers through Bailian, also known as Model Studio for international users.
Qwen3.7-Plus is designed to understand images and video alongside written prompts, distinguishing it from its text-only sibling, Qwen3.7-Max. This model focuses on visual understanding rather than generation, with Alibaba's generative AI models residing in separate families.
The new model introduces five key capabilities: deep reasoning, self-programming, tool invocation, verification and testing, and autonomous iteration. These features enable Qwen3.7-Plus to write and revise its own code, call external functions, and iteratively refine tasks, positioning it as an agent built for action rather than just answering.
Qwen3.7-Plus has shown strong performance in vision tasks, ranking #16 overall in Vision Arena, which places Alibaba as the #5 lab in vision. This indicates its proficiency in image-heavy applications such as OCR, chart reading, and video-frame analysis. Meanwhile, the text-only Qwen3.7-Max scored 56.6 on the Artificial Analysis Intelligence Index, making it the highest-ranked Chinese model at its release.
The Qwen3.7 series emphasizes an agentic focus, aiming for long-running tasks. The Bailian platform supports this with an Agentic RL mechanism for refining model accuracy through real-world feedback and built-in safety guardrails to ensure autonomous tools operate within preset limits.
This release from Alibaba's Qwen team marks a significant advancement in multimodal hybrid agent technology, offering a robust and versatile tool for developers and businesses.
Related articles
The AI world is getting ‘loopy’
AI models are taking a significant leap forward with the adoption of "agentic loops," where AI agents continuously prompt each other to improve code and solve complex problems. This approach, though potentially resource-intensive, promises to unlock new levels of autonomous problem-solving and efficiency in AI applications.
Codex-maxxing for long-running work
Codex is increasingly being used by organizations to support long-running projects that go beyond a single prompt. This whitepaper by Jason Liu offers practical strategies for leveraging Codex as a persistent workspace, managing complex workflows and sustaining progress.
Nobel laureate John Jumper is leaving DeepMind for rival Anthropic
Nobel laureate John Jumper is departing Google DeepMind to join its competitor, Anthropic, after dedicating nearly nine years to DeepMind, where he led the AlphaFold team. Jumper, who shared a Nobel Prize for his work on AlphaFold, expressed gratitude for his time at DeepMind while looking forward to new endeavors.