医生专用ChatGPT，估值120亿美元 又一家超级独角兽诞生。 Jay 发自 凹非寺 量子位 | 公众号 QbitAI 没想到，医疗场景ChatGPT生意能做这么大…… 最新消息，美国明星创业公司——医生版ChatGPT「OpenEvidence」，新融资将获投约 2.5亿美元，估值翻倍至 120亿美元 。 这意味着，Perplexity、Cursor等超级独角兽的牌桌上，将迎来一家医疗AI公司。 这并不令人惊讶，毕竟OpenEvidence几乎已经垄断了美国ToC医疗AI市场—— 日均处理临 床查询超6万次，45%的美国医生都是它的用户 。 Information消息，旨在打造医生专用ChatGPT的AI公司—— OpenEvidence，正筹集2.5 亿美元股权融资 。 如果交易完成，这家成立仅三年的初创公司估值将直接翻倍，来到120亿美元。 回看OpenEvidence的融资历史，也像火箭一样迅速攀升： 2025年2月，A轮融资7500万美元，估值10亿美元。 2025年7月，B轮融资2.1亿美元，估值翻了三倍多，来到35亿美元。 2025年10月，C轮融资2亿美元，估值进一步跃升至60亿美元。 股东 ...

医疗行业跨界洞察 本文来自微信公众号： 健闻咨询 ，作者：毛晓琼、李琳，原文标题：《问诊单量下降40%，基层医 生排不上班，线上问诊会是第一个被AI干掉的医疗赛道吗？》，题图来自：AI生成 做了近10年的线上问诊，舒浩没有想到，变化会来得如此之快。 舒浩是国内某头部三甲医院的一名内科高年资主治，从2016年开始接触线上问诊，几家主流的在线 问诊平台他都注册了个遍。从单量上来看，他在几个核心平台都算是靠前的，一个月平均能接500单 左右。 今年10月下旬，他所服务的某个主力平台发布了一款AI健康管家类产品。当时，这个偏行业侧的消 息并没有引起他的重视。直到11月过半，他发现自己在该平台上的线上问诊量缩水了40%左右。 以下文章来源于健闻咨询 ，作者健闻咨询 健闻咨询 . "其它平台的问诊量之前也有一些下降，大概在20%左右。"舒浩说。 在某互联网医疗公司任职的苏玉更早捕捉到了这种趋势，她所在的公司很早就围绕线上问诊做了独立 的业务布局，医生黏性强，在用户端的口碑也很好。 但从2024年下半年开始，她发现问诊平台的单量开始平稳下滑，"当时我们都不知道原因，后来接触 AI多了，大家才意识到，很多浅层的，和医疗健康相 ...

Core Insights - Hippocratic AI, an AI healthcare startup, has achieved a valuation of $3.5 billion (approximately 249 billion RMB) after raising $126 million in its latest funding round, marking a doubling of its valuation in less than a year [2][3] - The company focuses on developing non-diagnostic AI solutions to address global healthcare resource shortages, with applications in patient communication, information collection, and follow-up [2][12] Funding and Growth - Since its inception in May 2023, Hippocratic AI has completed multiple funding rounds, raising over $400 million in total [3] - The latest funding round (C round) took place in November 2025, raising $126 million, led by prominent investors including Andreessen Horowitz and General Catalyst [5] - The company has established partnerships with over 50 large healthcare systems and pharmaceutical companies across six countries, completing over 115 million clinical patient interactions without safety issues [7][18] Product and Technology - Hippocratic AI's core technology is a medical-specific large language model named Polaris, designed to ensure safety and practicality in patient interactions [12] - The company has launched a "Healthcare AI Agent App Store," allowing healthcare professionals to create and customize AI applications without programming knowledge, with over 1,000 non-diagnostic medical scenarios covered [17] - The AI agents have demonstrated high reliability, achieving a clinical evaluation accuracy rate of 99.38%, significantly surpassing the average of 81% for registered nurses in the U.S. [15] Market Position and Strategy - The company targets non-diagnostic patient communication tasks, such as pre-operative preparation and chronic disease management, effectively replacing traditional medical call centers [17] - The business model operates on a usage-based SaaS subscription, with costs approximately one-fourth of traditional human services, and high patient satisfaction ratings [17] - The global AI healthcare market is transitioning from concept to value realization, with a focus on addressing specific, measurable problems that meet regulatory requirements [21]

但从2024年下半年开始，她发现问诊平台的单量开始平稳下滑，"当时我们都不知道原因，后来接触AI 多了，大家才意识到，很多浅层的，和医疗健康相关的咨询已经开始被AI替代了。" "所有常规标准化的问诊，都会逐步被AI取代。"长期从事互联网医疗相关工作的资深人士徐晓向《健闻 咨询》表示，在他看来，这部分可被替代的问诊需求，在所有线上问诊场景里的占比接近80%。 做了近10年的线上问诊，舒浩没有想到，变化会来得如此之快。 舒浩是国内某头部三甲医院的一名内科高年资主治，从2016年开始接触线上问诊，几家主流的在线问诊 平台他都注册了个遍。从单量上来看，他在几个核心平台都算是靠前的，一个月平均能接500单左右。 今年10月下旬，他所服务的某个主力平台发布了一款AI健康管家类产品。当时，这个偏行业侧的消息 并没有引起他的重视。直到11月过半，他发现自己在该平台上的线上问诊量缩水了40%左右。 "其它平台的问诊量之前也有一些下降，大概在20%左右。"舒浩说。 在某互联网医疗公司任职的苏玉更早捕捉到了这种趋势，她所在的公司很早就围绕线上问诊做了独立的 业务布局，医生黏性强，在用户端的口碑也很好。 巧合的是，最近互联网医疗行业 ...

Core Insights - The IPO plans for two medical technology companies, Dingxiangyuan and Yilian, have stalled, with no specific updates available from either company [2][3] - Weiyi Holdings is actively pursuing an IPO in Hong Kong, having updated its prospectus in September [3][11] - The shift towards AI healthcare is evident as companies like Weiyi, Dingxiangyuan, and Yilian pivot from traditional internet healthcare to AI-driven solutions [4][7] Company Developments - Weiyi has transformed into an AI healthcare company, with over 90% of its revenue from AI medical services, totaling 28.41 billion yuan in the first half of 2025 [6][24] - Dingxiangyuan has launched AI tools like ClinMaster, focusing on clinical decision support, and has a significant user base of over 9 million registered professionals [6][15] - Yilian is developing a large medical model, MedGPT, and has been collaborating with companies for patient management solutions [20][29] Market Trends - The internet healthcare sector is struggling, with only drug sales proving profitable, as evidenced by the acquisition of Haodaifu Online by Ant Group [8][9] - Companies are increasingly adopting AI labels to attract investment, with recent IPOs in the AI healthcare space showing strong market interest [4][12] - The market for AI healthcare is hot, but investors remain cautious, focusing on business models, regulatory impacts, and unique competitive advantages [12][21] Business Models - Dingxiangyuan is shifting towards a consumer market, leveraging its large doctor user base to drive e-commerce without selling drugs [14][15] - Weiyi is focusing on a hybrid model that combines online and offline services, particularly in chronic disease management through partnerships with local healthcare systems [18][20] - Yilian is exploring two revenue streams: membership fees from individual users and expanding its corporate partnerships [28][29] Financial Performance - Weiyi's revenue grew by 69.4% year-on-year to 30.8 billion yuan in the first half of 2025, with a narrowing loss rate of 4.2% [24][25] - Dingxiangyuan's profitability remains uncertain, with its business model still needing to prove its viability in the consumer market [22][23] - Yilian's success hinges on establishing a robust technical barrier and acquiring quality medical data for its AI models [29][30]

Group 1: Meta's AI Integration - Meta will officially incorporate "AI-driven impact" into employee performance metrics starting in 2026, assessing how employees utilize AI to enhance work outcomes and team productivity [1] - The company has launched the "Level Up" game project and AI performance assistant tools this year to encourage employees to use the internal AI chatbot Metamate as much as possible [1] - Meta has begun allowing some job candidates to use AI assistants during coding interviews, believing this better represents a real development environment [1] Group 2: Google NotebookLM Features - Google NotebookLM introduced image data source functionality on November 15, enabling automatic OCR and semantic parsing, allowing users to retrieve content from images using natural language [2] - The underlying multimodal model can distinguish between handwritten and printed areas, extract table structures, and automatically link with existing text, audio, and video notes [2] - Within 48 hours of the feature launch, educational accounts uploaded over 500,000 pages of images, a 340% increase, with plans to integrate AR glasses for real-time "see and ask" capabilities next year [2] Group 3: Alibaba's Qianwen App Launch - Alibaba's Qianwen app public beta has launched, built on the Qwen3 model, providing an all-in-one entry point for users to experience a full suite of AI capabilities for free [3] - The application will gradually cover various life scenarios including office work, maps, health, and shopping, aiming to make AI a daily companion [3] - Qianwen will continue to evolve and integrate the latest Qwen models, currently available for search and download in major app stores in China [3] Group 4: Zhiyu GLM Coding Plan - Zhiyu has launched the "GLM Coding Plan·Special Edition" subscription package, offering a 50% discount for first-time buyers, with a minimum monthly cost of only 16 yuan [4] - Powered by the flagship model GLM-4.6, it ranked first globally in the LMArena evaluation alongside Claude Sonnet 4.5 and GPT-5, supporting 200K long context [4] - The model is officially compatible with over 10 mainstream AI programming tools, with several US tech companies like Cerebras and Vercel adopting GLM-4.6 [4] Group 5: Xiaomi's Miloco Solution - Xiaomi has launched its first "large model + smart home" solution, Miloco, using the Mijia camera as a visual information source, with the self-developed large language model MiMo-VL-Miloco-7B at its core, and the framework is open-sourced [5] - Users can communicate with the smart home system through natural language, allowing the system to automatically fulfill various smart needs and rules while ensuring privacy through visual data understanding [5] - Xiaomi's AIoT platform has connected nearly 1 billion IoT devices, and Miloco achieves interoperability between the Mijia ecosystem and Home Assistant ecosystem through standardized MCP protocols, supporting third-party IoT platform integration [5] Group 6: MiroMind's MiroThinker v1.0 - MiroMind has officially launched the open-source intelligent agent base model MiroThinker v1.0, introducing a new dimension of "deep interaction scaling," supporting 256K context and 600 tool calls [6] - In the BrowseComp test, it achieved an accuracy rate of 47.1%, nearing OpenAI DeepResearch's 51.5%, while surpassing DeepSeek-v3.2 by 7.7 percentage points in Chinese tasks [6] - The model adopts a fully open-source architecture, providing all model weights, toolchains, and interaction frameworks, with the 72B version approaching or even surpassing OpenAI DeepResearch, promoting intelligent agents from passive execution to active learning evolution [6] Group 7: MedGPT's Clinical Success - The core model of Future Doctor AI Studio, MedGPT, has outperformed GPT-5 and other leading international models in a multi-model practical evaluation conducted by 32 top domestic clinical experts, achieving the global first in clinical safety and effectiveness assessment [7] - It has launched two products: a clinical decision AI assistant and a patient follow-up AI assistant, providing safe and effective decision support during diagnosis and supporting patient follow-up for chronic disease management [7] - MedGPT has been adopted by dozens of national discipline leaders for daily use and is recognized by experts as the "best practice" for AI empowering grassroots healthcare, aligning with the National Health Commission's guidelines for promoting and regulating AI in healthcare [7] Group 8: Li Feifei on AGI - Li Feifei stated in an interview that AGI is "more of a marketing term than a scientific term," emphasizing that the current AI's biggest shortcoming is the lack of spatial intelligence, which allows humans to navigate and manipulate in a three-dimensional world [8] - She outlined three core capabilities of world models: generative, multimodal, and interactive, arguing that relying solely on data and computing power will not lead to the maturity of robots, which are physical systems needing bodies and application scenarios [8] - The first large-scale world model product, Marble, released by World Labs, has been widely applied in film production, game development, scientific research, and robot training, reducing creation time by 40 times [8]

Core Viewpoint - The article emphasizes the importance of AI in enhancing the efficiency and safety of grassroots healthcare, particularly through the "Future Doctor AI Studio," which has been recognized for its clinical decision-making and patient follow-up capabilities [4][72]. Group 1: Policy and Implementation - The National Health Commission has prioritized "AI + grassroots application" as a key direction in its recent policy, aiming for comprehensive coverage of intelligent auxiliary applications in grassroots diagnosis and treatment by 2030 [4][72]. - The implementation of AI in healthcare is seen as a response to the increasing workload and complexity faced by grassroots doctors, who often struggle with time constraints and patient management [3][5]. Group 2: AI Capabilities and Evaluation - The "Future Doctor AI Studio" utilizes a model called MedGPT, which has been evaluated and found to outperform leading international models like OpenAI's GPT-5 in terms of safety and effectiveness in clinical settings [13][72]. - A clinical evaluation involving 32 top domestic experts highlighted that MedGPT achieved the highest scores in safety and effectiveness, significantly surpassing other models by 15.3% [13][17]. Group 3: Practical Applications - The AI system is designed to assist doctors in two critical areas: clinical decision-making during patient consultations and managing follow-up care for chronic disease patients [21][38]. - The clinical decision-making AI assistant helps doctors quickly identify risks and necessary actions in high-pressure situations, while the patient follow-up AI assistant monitors patients post-consultation, ensuring ongoing care and timely interventions [24][43]. Group 4: User Feedback and Adoption - Feedback from healthcare professionals indicates that the "Future Doctor AI Studio" effectively reduces anxiety and enhances decision-making confidence among doctors, making it a trusted tool in clinical practice [34][66]. - The AI's design focuses on usability and practical support rather than flashy features, which has led to its rapid adoption among healthcare providers [51][67].

Government Initiatives - The State Council issued opinions to accelerate the cultivation and large-scale application of new scenarios, emphasizing the importance of connecting technology and industry for promoting commercialization of new technologies and products [1] - The Ministry of Industry and Information Technology (MIIT) announced plans to enhance the systematic layout and high-level construction of manufacturing pilot platforms, aiming to establish a modern pilot platform system by the end of 2027 [1] - MIIT initiated the 2025 AI industry and new industrialization innovation tasks, focusing on key areas such as AI infrastructure and "AI + manufacturing" to foster technological innovation and application [1] Local Actions - Jiangsu Province is promoting a new paradigm for immersive sports event consumption by integrating technologies like AI, big data, and digital twins into sports event activities [4] - Guangdong Province is collecting quality projects for smart tourism, focusing on new information technologies such as cloud computing, AI, and the metaverse to enhance tourism experiences [5][6] Industry Developments - Neuralink has implanted devices in 12 individuals, enabling them to regain abilities and pursue careers, showcasing advancements in brain-machine interface technology [9] - A new clinical safety-effectiveness dual-track benchmark was established by 32 clinical experts, with a Chinese model, MedGPT, achieving the highest score, indicating significant progress in domestic medical AI capabilities [10] Insights from Experts - Academician Zhang Rong highlighted the synergy between AI and Micro-LED technology, predicting breakthroughs in both fields due to their mutual reinforcement [11] - The next generation of remote sensing systems is evolving into intelligent agents capable of understanding and reasoning, driven by advancements in AI and remote sensing technologies [13] Technological Innovations - A new AI model developed by DeepMind successfully predicted hurricane paths and intensity, demonstrating the potential of AI in disaster management [15] - Huawei Cloud introduced a versatile intelligent platform aimed at addressing challenges in AI deployment, focusing on reducing development barriers and enhancing edge capabilities [16]

Core Viewpoint - The article emphasizes the importance of AI in grassroots healthcare, highlighting the need for safety, effectiveness, and human-AI collaboration as essential criteria for successful implementation [2][4][44]. Policy and Market Context - On November 4, the National Health Commission issued a document outlining the core goal for the next five years: "AI + grassroots application," placing it at the forefront of the eight key directions for "AI + healthcare" [4]. - The document aims for "intelligent auxiliary applications in grassroots diagnosis and treatment to achieve basic coverage by 2030" [5]. Current Challenges - Despite the policy push, there is a significant gap in AI adoption at the grassroots level, with over 80% of grassroots doctors not using AI, and those who do often rely on generic models that lack precision [7]. - The article notes a "reverse situation" where major hospitals are rapidly adopting AI, while grassroots healthcare remains largely untouched by the AI wave [7]. AI Product Features - The "Future Doctor AI Studio" is presented as a reliable tool that aligns with the policy blueprint, focusing on safety and effectiveness [9]. - MedGPT, the underlying model of the Future Doctor AI Studio, has been rigorously tested for safety and effectiveness, outperforming five major global models in clinical scenarios [12][14]. Safety and Effectiveness - MedGPT achieved the highest scores in safety (0.912) and effectiveness (0.861) during evaluations, significantly surpassing other models [17]. - The article stresses that true medical AI must prioritize safety and effectiveness, with clinical value as the benchmark for technological iterations [11][13]. Human-AI Collaboration - The article highlights the importance of human-AI collaboration, stating that AI should serve as a "super assistant" to doctors, enhancing their capabilities rather than replacing them [39][40]. - The Future Doctor AI Studio's clinical decision-making assistant is designed to support grassroots doctors by providing structured decision reports based on high-level medical evidence [22][25]. Clinical Decision Support - The clinical decision AI assistant can generate comprehensive decision reports for complex cases, demonstrating expert-level reasoning and reliable decision-making [23][30]. - Recent evaluations showed that the assistant outperformed competitors in various clinical scenarios, confirming its effectiveness in real-world applications [27]. Patient Follow-Up - The patient follow-up AI assistant addresses the critical "last mile" of healthcare, ensuring continuous patient management and communication [32][35]. - It automates follow-up tasks, provides personalized health management plans, and alerts doctors to high-risk signals, thereby enhancing patient care [36][38]. Conclusion - The article concludes that the integration of AI in grassroots healthcare represents a best practice for empowering medical professionals and improving patient outcomes, with a strong emphasis on safety, effectiveness, and collaboration [44].

Core Viewpoint - The article discusses the gap between the real capabilities of medical AI and clinical expectations, highlighting the need for a new evaluation standard for medical AI that reflects its clinical applicability and safety [1][10][17]. Group 1: Medical AI's Current Challenges - Many AI models that perform well in standardized exams often reveal issues such as reasoning errors, misdiagnosis, and inappropriate treatment plans in real clinical settings [2][9]. - The recent update from OpenAI prohibits ChatGPT from assisting in medical diagnostics, indicating a cautious approach towards AI's involvement in serious medical fields [2][10]. Group 2: New Evaluation Standards - A new evaluation standard, the Clinical Safety-Effectiveness Dual-Track Benchmark (CSEDB), has been developed by top Chinese clinical experts to assess the clinical applicability of medical AI [5][10]. - This new standard introduces a dual evaluation system focusing on both safety and effectiveness, moving beyond traditional exam score metrics [11][12]. Group 3: MedGPT's Performance - MedGPT, a model developed by a Chinese company, achieved the highest score of 0.895 in the CSEDB evaluation, outperforming other models by over 15 percentage points [19][22]. - MedGPT is the only model to score higher in safety than in effectiveness, demonstrating a cautious approach in clinical scenarios [24][26]. Group 4: Future of Medical AI - The future of medical AI lies in its ability to replicate the expertise of top clinicians, creating new medical resources and serving as a reliable assistant for healthcare professionals [34][46]. - The "Future Doctor" platform aims to scale the clinical experience and decision-making capabilities of expert doctors through AI, ensuring that all patient interactions are handled by real doctors [41][45]. Group 5: Industry Impact - The establishment of the CSEDB standard represents a significant step towards a more mature medical AI industry, allowing for better evaluation and optimization of AI models [54][55]. - The evolution of medical AI from merely simulating doctor responses to actively participating in clinical reasoning marks a pivotal moment in the industry [56][57].