牛透社 10 月 15 日消息，阿里巴巴人工智能事业部发布 Qwen3-VL 视觉语言模型系列的轻量化版本，推出 4 亿（4B）与 8 亿（8B）参数变体。该产品针对 STEM 推理、视觉问答（VQA）、光学字符识别（OCR）等核心多模态能力进行优化，基准测试显示其多项指标超越 Gemini2.5 Flash Lite 与 GPT-5 Nano 等竞品，部分表现看齐 Qwen2.5-VL-72B ...
Source Link牛透社 10 月 15 日消息，阿里巴巴人工智能事业部发布 Qwen3-VL 视觉语言模型系列的轻量化版本，推出 4 亿（4B）与 8 亿（8B）参数变体。该产品针对 STEM 推理、视觉问答（VQA）、光学字符识别（OCR）等核心多模态能力进行优化，基准测试显示其多项指标超越 Gemini2.5 Flash Lite 与 GPT-5 Nano 等竞品，部分表现看齐 Qwen2.5-VL-72B ...
Source LinkDisclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.