Core Insights - China's open-source AI ecosystem is rapidly evolving, showcasing technological confidence and creating a path for global collaboration, contrasting with the closed-source approach prevalent in the U.S. [1][6][8] Group 1: Open-Source Development in China - DeepSeek and other foundational models like Alibaba's Qwen are driving the advancement of China's open-source capabilities, leading to the emergence of smaller, more powerful vertical models from various SMEs [1][4] - The launch of models like Skywork-OR1 by Kunlun Wanwei demonstrates that even companies with limited funding can achieve state-of-the-art (SOTA) performance by leveraging existing open-source models [4][5] - The rapid iteration of large models in China, such as Alibaba's Qwen2.5-VL and the multi-modal models from Jiepu, indicates a thriving open-source ecosystem [5][6] Group 2: Comparison with U.S. AI Strategy - The U.S. AI industry remains predominantly closed-source, driven by major tech companies and venture capitalists seeking high returns, which fosters a monopolistic environment [6][8] - OpenAI's shift to a closed-source model, particularly after its partnership with Microsoft, highlights the commercial motivations behind this strategy [7][8] - In contrast, China's top-down approach emphasizes open-source development as a means to enhance technological equity and industry collaboration [8][9] Group 3: Economic and Social Implications - The Chinese government has actively supported open-source initiatives, recognizing their potential to lower technological barriers and promote economic integration [8][9] - Investments in open-source projects, such as the Z Fund's commitment to support AI open-source communities, reflect a broader strategy to foster innovation [9][10] - The open-source movement in China is not only about providing free products but also about enabling developers to build upon existing technologies, thus accelerating progress [5][10] Group 4: Practical Applications and Success Stories - Open-source models are being successfully implemented in various industrial applications, such as predictive maintenance in manufacturing and environmental conservation efforts [13][14] - Companies like Baosteel and Zhongmei Kegong are utilizing open-source models to enhance operational efficiency and reduce costs [13][14] - The collaborative nature of open-source development allows for broader participation in AI projects, benefiting both commercial and non-profit sectors [14][15] Group 5: Future Outlook - China's open-source AI landscape is transitioning from a phase of "technological following" to "ecosystem leadership," reshaping the global AI landscape [18][20] - The focus is shifting from mere parameter competition to the deep integration of AI technology with the real economy, indicating a strategic evolution in the industry [18][20]

Core Insights - The "Zhibao Black Soil" platform, driven by DeepSeek technology, is China's first intelligent configuration platform for black soil erosion control, currently in trial operation [1][3] - This platform acts as a "digital black soil expert," utilizing deep learning to analyze decades of soil research, meteorological data, and agricultural practices to create a comprehensive digital archive of black soil characteristics [1][3] - The platform can quickly assess black soil erosion features and generate customized control solutions based on local environmental conditions, significantly reducing the time required for such assessments from months to half an hour [3] Industry Impact - The platform is currently being tested in demonstration bases in Heilongjiang and Jilin, with plans to cover major black soil areas in Northeast China by 2026, contributing to the protection of deteriorating black soil [3] - The initiative is part of China's "14th Five-Year" national key research and development plan, aiming to enhance the sustainable use of black soil resources and promote innovation in soil conservation [3]

Group 1 - The core viewpoint emphasizes the importance of applications over models in the AI landscape, as articulated by Baidu's founder, Li Yanhong, during the Create2025 Baidu AI Developer Conference [2] - Baidu launched a "nine-piece set" of tools and models aimed at reducing costs and enhancing capabilities for developers, including two new models with up to 80% price reduction [3] - The rapid iteration of models raises questions about the longevity of application value, but Li Yanhong asserts that finding the right scenarios and models will ensure applications remain relevant [2][3] Group 2 - Baidu introduced two new models, Wenxin Model X1 Turbo and 4.5 Turbo, which are multi-modal and strong reasoning models, indicating a shift towards multi-modal models as the future standard [3] - The company is also focusing on no-code programming tools like Miaoda and the general-purpose intelligent agent "Xinxiang," which can generate applications and provide comprehensive solutions to complex user problems [4] - The industry is witnessing a rapid evolution in application development, with major tech companies like Alibaba and Tencent also launching competitive products and services to support developers [4]

Core Insights - Baidu's founder, Li Yanhong, emphasized that DeepSeek is not a panacea and highlighted its current limitations, particularly in processing multimedia content [2][3] - DeepSeek achieved significant success by becoming the fastest application to surpass 30 million daily active users and topping the App Store charts in multiple regions, including the US [2] - The AI industry is witnessing a trend towards multimodal models, which are expected to become standard in future foundational models [6] Industry Overview - The training costs for mainstream large models in China typically range from tens of millions to hundreds of millions of dollars, with major players like Baidu, Alibaba, and Tencent investing over $200 million [4] - Startups like Kimi and DeepSeek have managed to reduce training costs to between $30 million and $60 million through technological optimizations [4] - Revenue from multimodal large models in China is concentrated among leading companies, with Alibaba Cloud generating over 110 billion yuan, accounting for about 15% of its group revenue [5] Application Insights - Li Yanhong stressed the importance of applications over models and chips, asserting that the true value lies in the applications that utilize these technologies [6] - Despite DeepSeek's shortcomings, the focus remains on finding the right scenarios and models to create lasting applications [6]

Group 1 - The core viewpoint of the article highlights BMW's strategic move to deepen its AI ecosystem in China by integrating with DeepSeek, following its collaboration with Alibaba on AI large language models [1] - BMW China announced that the DeepSeek functionality will be applied to new generation models produced in China, enhancing the human-machine interaction experience centered around the BMW Intelligent Personal Assistant [1] - Starting from the third quarter of this year, the DeepSeek feature will be implemented in several new cars sold in China that are equipped with the 9th generation BMW operating system [1]

图片来源：维基百科 百度创始人李彦宏在周五举行的开发者大会上发表重要演讲，对中国AI领域的新锐力量提出建设性意见。他指出，尽管DeepSeek公司开发的纯文本模型 R1在开发者社区获得广泛认可，但随着行业技术发展，单一文本处理模型的市场需求正在"逐步缩减"。 在发布新一代多模态模型Ernie 4.5 Turbo和X1 Turbo时，李彦宏强调，百度最新研发的模型不仅具备文本处理能力，还能高效整合音频、图像和视频等多 种数据形态。他进一步分析指出，与国产同类产品相比，纯文本模型在运行效率、成本控制以及避免"幻觉"输出等方面存在明显劣势。 百度在大会上重点展示了多模态AI技术的商业应用前景，包括升级版的数字人平台。该解决方案能够为企业客户打造高度仿真的虚拟主播，有效支持直 播电商和产品营销等场景。 值得注意的是，尽管指出竞争对手的技术局限性，百度仍采取开放态度，将DeepSeek的模型整合至自身生态系统。自今年1月DeepSeek凭借R1推理模型崭 露头角后，百度已将其接入千帆企业平台，并成功应用于地图和搜索等核心产品线。 Forrester Research副总裁戴查理分析认为，百度此次技术发布将"显著推动中国 ...

文｜周鑫雨 编辑｜苏建勋 如今，模型的发展已经成了应用开发者的一把双刃剑。一方面，模型能力的增强，给场景落地带来更多的可能；另一方面，应用的能力，也随时可能被模 型能力本身淘汰。 什么样的应用不会过时？ 2025年4月25日，百度Create大会上，百度集团创始人、董事长兼CEO李彦宏的答案是："你只要找对场景，选对基础模型，有时候可能还要学一点调模型 的方法，那么在这个基础上做出来的应用，它是不会过时的，应用才是真正创造价值的。" 在应用层，李彦宏认为2025年以来最令人激动的突破性的应用，是数字人和Agent。 卖数字人，是百度布局电商业务的重要一环。在会上，李彦宏发布了最新的数字人能力：高说服力数字人。除了更加拟人，"高说服力数字人"的特点，是 可以感知直播间环境，并智能做出发红包、换品等决策。 百度在应用层的另一个布局重点，是Agent。 一年前，李彦宏曾表示，AI Coding是他最看好的应用方向。目前，百度在代码智能体领域已经做了较为完整的布局，有面向专业工程师的编程工具 Comate，面向普通人的无代码编程工具"秒哒"。 百度的"App版Manus"发布了 2025年3月6日发布的Agent应用 ...

文 丨 新浪科技 周文猛 4月25日，于樱花之都武汉市举办的Create 2025百度AI开发者大会上，百度创始人、董事长兼CEO李彦 宏发表了他2025年的首场个人演讲。 长达一个小时的分享中，"应用才是王者""DeepSeek不是万能的""成本下降""智能体协同""MCP"等词汇 高频出现，串联起了他对于AI应用以及模型发展的思考以及洞见，同时也巧妙地给自家产品进行了带 货。 以下，新浪科技梳理了本场大会李彦宏重点提及的6大核心观点，供读者参考。 "模型会有很多，但应用才是王者" 过去一年，伴随模型降价以及DeepSeek横空出世并开源，越来越多人加入开发者队伍。但是，模型的 快速迭代，也正让开发者们变得焦虑，担心自己的模型选型出错导致开发的应用很快过时，进而失去价 值。 对此，李彦宏深表认同。不过他也进一步指出，"这么多日益强大的模型能力，其实也为开发者提供了 更多的选择，打开了更多的可能性，你只要找对场景，选对基础模型，会调模型，做出来的应用就不会 过时。" 在李彦宏看来，应用才是真正创造价值的，没有应用，芯片、模型都没有价值。"模型会有很多，但未 来真正统治这个世界的是应用，应用才是王者。" 在他看 ...

如何证明自家的模型是行业一流水平？跟DeepSeek比一比。 随着DeepSeek爆火，2025年的大模型市场，DeepSeek就像手机圈的iPhone一样，成了各家模型厂商发布新品时的最新对标对 象。 除了模型能力上号称超过DeepSeek之外，上述对标更直观的表现在价格上。 在今天百度文心大模型4.5 Turbo、文心大模型X1 Turbo发布时，百度创始人李彦宏就吐槽DeepSeek也不是万能的。除了能力上 只能处理文本，还不能理解和生成图片、音频、视频等多媒体内容、幻觉率高外，李彦宏认为使用DeepSeek更大的问题是慢和 贵，"中国市场上绝大多数的大模型API调用价格都比DeepSeek满血版要低，而且速度也更快。" 当然，其中也包括最新发布的文心4.5 Turbo和文心X1 Turbo。官方数据显示，文心4.5 Turbo每百万token的输入价格为0.8元，输 出价格3.2元。 作为对比，即便在打五折后的优惠时段（00:30-8:30），DeepSeek V3每百万token的输入价格也达到1元，输出价格4元。 推理模型上，百度模型的性价比则进一步凸显：文心大模型X1 Turbo每百万token ...

作者 | 梁昌均 "模型的世界，应用的天下。"这是4月25日在武汉举办的百度Create开发者大会的主题。 出品 | 搜狐科技 百度创始人、董事长&CEO李彦宏在会上强调，模型会有很多，但真正统治这个世界的是应用，应用才是王者。 "AI的发展日新月异，新技术层出不穷，令人眼花缭乱，甚至不知所措。"李彦宏表示，过去一年，模型的发布越来越频繁，DeepSeek横空出世，MCP逐渐 被接受为行业标准，代码智能体和通用智能体被广泛关注，越来越多的人加入到开发者的队伍。 此次大会，李彦宏推出文心最新版本模型，在实现性能提升的同时，价格低于DeepSeek。同时，百度还公布了数字人、智能体、芯片等方面的应用进展。 李彦宏坚定看好AI应用的发展，认为随着模型能力的增强，大模型和应用场景结合的机会会越来越多。这些AI应用不但不会被大模型的能力所覆盖，反倒 能够渗透到更多的场景，是属于开发者的真机会，"千行百业遍地是黄金"。 李彦宏还提到了DeepSeek，在其发布以后，百度第一时间在百度智能云进行了部署，文小言、百度搜索、百度地图等也都接入。 不过，李彦宏表示，DeepSeek也不是万能的，它只能处理单一的文本，还不能理解声 ...