阿里开源 Qwen-Image-2.1：70 亿参数图像模型称超越闭源
原标题：Alibaba's open-weight Qwen-Image-2.1 claims to beat closed models in image generation with just 7 billion parameters
AI 摘要
阿里巴巴 Qwen AI 团队发布开源权重图像生成与编辑模型 Qwen-Image-2.1，其视觉生成组件仅 70 亿参数，团队称在自家基准上超过多数闭源模型，但独立评测尚未出炉。该模型可在 3090 等消费级 GPU 上运行，原生支持 RGBA 透明图像生成与编辑，最多可同时处理十张参考图，并通过架构调整与 KV 缓存复用加速推理。模型已在 Hugging Face、GitHub 和 Model Scope 发布，研究许可禁止商业使用，企业需另行申请授权。
正文节选
Alibaba's open-weight Qwen-Image-2.1 claims to beat closed models in image generation with just 7 billion parameters Alibaba's Qwen AI team has released Qwen-Image-2.1, an open-weight model for image generation and editing. Its visual generation component has just 7 billion parameters yet beats most closed models on Qwen's own benchmark, the team claims, though independent benchmarks are still pending. It runs on capable consumer GPUs like a 3090. The model natively generates and edits transpare