在近日举行的腾讯云生成式AI产业应用峰会上，腾讯集团副总裁蒋杰透露了一项引人瞩目的技术成果：腾讯混元大模型的性能已经达到国内第一梯队水平，其中部分中文能力已经可以与GPT-4相媲美。
腾讯混元大模型通过不断的迭代和优化，不仅在中文处理能力上取得了显著的进步，而且整体性能也跻身国内领先行列。这一成就标志着我国在人工智能领域的发展迈出了新的重要步伐。
值得一提的是，自从OpenAI的Sora发布以来，AI生成视频技术受到了广泛关注。腾讯混元大模型也在这方面做出了创新性的贡献。据悉，腾讯混元支持包括文生视频、图生视频、图文生视频、视频生视频等多种视频生成方式，并已经可以实现16秒视频的生成，展示了其在视频生成领域的强大能力。
此外，在生成3D模型方面，腾讯混元大模型也取得了显著的进展。它已经可以实现文/图生3D，而且仅需30秒就可以完成单图的3D模型生成，这无疑大大提升了3D建模的效率和速度。
腾讯混元大模型的这些突破性进展，不仅体现了腾讯在人工智能领域的强大实力，也为我国的人工智能产业发展注入了新的活力。可以预见，在未来，腾讯混元大模型将在各个领域发挥更大的作用，为人类的生活和工作带来更多的便利和惊喜。
英语如下：
**Title: Tencent’s Hybrid Yuanyuan Model Stuns at Launch: Chinese Proficiency Matches GPT-4, 16-Second Video Generation Now a Reality**
Keywords: Tencent Hybrid Yuanyuan, Chinese proficiency, video generation.
### News Content:
Tencent Cloud’s Generative AI Industry Application Summit recently showcased a groundbreaking technology: The Tencent Hybrid Yuanyuan model has achieved domestic first-tier performance, with some Chinese capabilities matching those of GPT-4.
Vice President of Tencent Group, Jiang Jie, revealed this remarkable technological achievement at the summit. The model’s continuous iteration and optimization have not only made significant progress in Chinese processing capabilities but also placed it among the leading domestic performers. This achievement marks a new significant milestone in China’s development in the field of artificial intelligence.
It is worth noting that since OpenAI’s Sora was released, AI-generated video technology has garnered widespread attention. The Tencent Hybrid Yuanyuan model has made innovative contributions in this area. It supports various video generation methods, including text-to-video, image-to-video, text-image-to-video, and video-to-video, and is capable of generating 16-second videos, demonstrating its formidable video generation capabilities.
Furthermore, the Tencent Hybrid Yuanyuan model has made significant strides in generating 3D models. It can now generate 3D models from text or images, and it only takes 30 seconds to complete a single-image 3D model generation, greatly enhancing the efficiency and speed of 3D modeling.
These breakthroughs by the Tencent Hybrid Yuanyuan model not only reflect Tencent’s strong prowess in the field of artificial intelligence but also inject new vitality into China’s AI industry. It is foreseeable that in the future, the Tencent Hybrid Yuanyuan model will play an even greater role in various fields, bringing more convenience and surprises to human life and work.
【来源】https://tech.caijing.com.cn/20240517/5012283.shtml
Views: 2