本文为IPO早知道原创 作者｜MD 微信公众号｜ipozaozhidao A股迎来了多日上涨，商业航天概念和硬科技驱动"功不可没"。 新鼎资本投资解决「卡脖子」问题的公司，也看好"马斯克的投资方向"。 2025年，国内的股权投资机构也迎来了久违的"市场回暖"，IPO早知道统计，2025年共有超过千家 机构收获了IPO项目。其中，新鼎资本收获了四个IPO项目，位列「 2025年度IPO最佳投资机构 TOP100 」的第 48位，项目 其中包括希迪智驾这样新鼎资本重仓的项目，累计投入 3亿多元。 面向未来，新鼎资本董事长张驰对 IPO早知道表示，要全力抓住接下来的商业航天浪潮。 新鼎资本在五年前开始布局商业航天。 " 国内成功发射民用火箭的六家公司中，我们投了 三 家， 分别为蓝箭航天、星际荣耀和中科宇航 。 "这背 后的原因实际上是看到了S pace X的成功，张驰 也并不回避这一点。 此前新鼎资本也曾布局新能源、自动驾驶等领域，在这一轮投资浪潮结束后，开始布局 AI应用层和 芯片供应链上"卡脖子"的环节。 但在回顾过去一年的项目时，张驰也坦言，自己并没有完全看懂 GPU公司。" 摩尔线程 ，我们实际 上投得不 ...

IPO前估值为4.85亿美元。 本文为IPO早知道原创 作者｜罗宾 微信公众号｜ipozaozhidao 来源：公司招股书 亦诺微核心产品 MVR-T3011是一种基于HSV-1的新型溶瘤免疫疗法，其独特之处在于将强效的肿 瘤裂解与表达抗 PD-(L)1 抗体和 IL-12 相结合，靶向治疗全谱系膀胱癌及包含头颈鳞癌的更广泛 实体瘤类型。 公司已为 MVR-T3011 设计了一套结构化的 开发路线图 。 在膀胱癌领域，公司采取分阶段策略覆 盖全谱系疾病。针对高危卡介苗（ BCG）无应答型非肌层浸润性膀胱癌（NMIBC），关键的II期临 床试验（NCT06971614）已于2025年6月在美国完成首例患者给药，正在评估其作为二线治疗方 案的潜力。去年12月，公司在2025年美国泌尿肿瘤学会年会上发布了MVR-T3011治疗BCG无应答 NMIBC的IIa期研究的阶段性疗效数据（截至2025年9月），数据显示出令人鼓舞的疗效和良好的安 全性。 同时，公司积极探索 MVR-T3011在卡介苗未治型NMIBC和肌层浸润性膀胱癌（MIBC）中的治疗 潜力，目前处于早期研究阶段。 在 MVR-T3011 头颈鳞癌临床进展 ...

Core Viewpoint - The article highlights the rapid globalization strategy of INMO, a leading player in the AI+AR smart glasses sector, which has successfully completed multiple funding rounds totaling nearly 500 million yuan within a year, indicating strong capital interest in the company [2][3]. Group 1: Financing and Market Position - INMO has completed three rounds of financing (B2, B3, and C1) within a year, with a total amount close to 500 million yuan, making it one of the most capital-attractive companies in the AI+AR smart glasses market [2]. - As a pioneer in the smart glasses industry, INMO launched the first wireless all-in-one AR smart glasses, INMO AIR, in 2022, leading the industry into a "full-color lightweight" phase [3]. - The introduction of INMO GO, the world's first AI glasses, in 2023 marked the entry of the industry into the "AI era," transforming futuristic concepts into reality [3]. Group 2: Product Innovation and Market Response - INMO GO2 has become a breakout product, validating the company's product strategy and attracting significant consumer interest [4]. - The upcoming flagship models, INMO AIR3 and INMO GO3, are set to launch in 2025, with INMO AIR3 being the world's first mass-produced 1080P all-in-one AI+AR glasses, which sold out immediately upon release [4]. - INMO GO3 has achieved over 20,000 orders within three days of its launch, showcasing its strong market performance across various metrics [7]. Group 3: Technological Advancements - INMO is the only smart glasses company committed to the "next-generation phone" mission, leveraging self-developed core technologies and partnerships to create AI glasses with proactive service capabilities [6]. - The INMO GO3 features advancements such as context understanding and real-time user assistance, enhancing user interaction and experience [6]. - INMO AIR3 utilizes AutoGLM technology to optimize on-device inference efficiency, enabling practical functionalities like voice-controlled operations and automated task execution [6]. Group 4: Brand Influence and Global Expansion - INMO has increased its brand visibility through offline advertising in key urban areas and endorsements from celebrities, transitioning smart glasses from a tech product to a lifestyle symbol [8]. - The INMO GO series has gained popularity on social media and entertainment platforms, further solidifying its market presence [8]. - Following a successful showcase at CES, INMO plans to expand its global footprint with new product versions and a comprehensive online and offline distribution network [10][11].

Core Insights - Qingtian Rental has successfully completed seed round financing, led by Hillhouse Ventures, with participation from several other investors, aiming to expand its market presence and enhance its technology and service systems in the Robot as a Service (RaaS) sector [3][6] - The platform has achieved significant user engagement, with over 200,000 registered users and an average of more than 200 rental orders per day within three weeks of launch, indicating a strong growth trajectory [4][6] - Qingtian Rental's innovative model transforms high-cost robotic products into on-demand, standardized service capabilities, addressing the industry's challenges of high application costs and fragmented service chains [3][4] Company Strategy and Leadership - The leadership team includes experienced professionals from various sectors, with a clear "1234 strategy" aimed at connecting with over 10 robot manufacturers, developing 200 premium service providers, gathering 3,000 content creators, and serving 400,000 rental customers by 2026 [6][8] - The CEO emphasizes the importance of not just being a time leader but also taking responsibility for industry standards and ecological impact, aiming to drive the large-scale application of robots [6][8] Industry Context and Challenges - The robotics industry is transitioning from technology validation to commercial implementation, with the main challenge being the reduction of usage barriers and improvement of delivery efficiency [7][8] - Qingtian Rental's platform approach is seen as a solution to the "last mile" problem in robot commercialization, facilitating the integration of robots into real-world business environments [6][8] - The focus on high-frequency interaction and strong experiential attributes in sectors like offline consumption and tourism is identified as a key area for the scalable deployment of robotic services [8][9]

Noumena不是传统的Agency，是一家AI Agent的公司。 本文为IPO早知道原创 作者｜ Stone Jin 微信公众号｜ipozaozhidao 据IPO早知道消息，物自体科技（Noumena）日前完成数千万元Pre-A轮融资，本轮融资投资方包 括狮城资本、百度战投以及老股东靖亚资本。 本轮资金将主要用于AI Agent产品迭代，包括营销垂类大模型的训练，面向AI决策的上下文图谱 （Decision Context Graph），多智能体（Multi-Agent）协作架构的技术研发，以及在美妆、时 尚等头部行业的深度拓展，并加速全球出海业务的布局。 成立于2024年的Noumena由前SAP全球副总裁、前第四范式总裁裴沵思创立，团队核心成员约40 人，其中超过60%是AI大模型、算法和工程研发，核心技术成员拥有清华、上海交大、港科大等顶 尖高校博士背景，并曾经在阿里巴巴、字节跳动、第四范式等领先的AI公司从事大模型等核心产品研 发工作。同时，团队汇聚了来自小红书、抖音、TikTok、SAP等企业的核心业务运营人员，具备丰 富品牌增长实战经验， 形成了在 AI 技术研发与营销业务两端均具备深度经验 ...

引领亚太具身智能新浪潮。 本文为IPO早知道原创 作者｜ Stone Jin 微信公众号｜ipozaozhidao 据IPO早知道消息，智元机器人海外首家机器人体验馆于当地时间2026年1月13日在马来西亚正式 开业。这座体验馆位于马来西亚雪兰莪州莎阿南i-City 内，是马来西亚第一个AI机器人体验馆，也 是智元在海外的首家体验馆。 这座由智元机器人和i-City共同打造的体验馆占地约2000平米，作为一个兼具公众体验、文化娱乐 及沉浸式AI应用的平台， AI World体验馆通过大型的沉浸式展示空间，集中呈现人工智能和人形机 器人技术在未来住宅、商业、健康等场景的应用。 据 此 前 全 球 权 威 市 场 研 究 机 构 Omdia 发 布 的 《 通 用 具 身 机 器 人 市 场 雷 达 》 报 告 显 示 ， 智 元 以 5100+台出货量位居全球人形机器人出货量第一。此次活动标志着智元作为头部品牌率先进入马来 西亚与亚太市场，并将其作为全球化战略布局的关键一步。 马来西亚科技与创新部长郑立慷在仪式上致辞，并宣布体验馆开幕。他表示，AI World体验馆是马 来西亚实施国家人工智能行动计划、引领区域 ...

据IPO早知道消息，浙江大学与康迪科技近日在紫金港校区签署战略合作协议，共同揭牌成立"浙江 大学-康迪智能机器人联合研发中心"。 双方的合作始于共同认知：未来的竞争，是体系与体系的竞争，是生态与生态的对话。基于此，双方 将各自发挥优势，互相赋能，构建体系。 在国际市场拼杀多年的康迪科技，以其全面丰富的UTV（多功能全地形车）与电动非公路车体系切 入了北美高端市场，出口份额稳居国内榜首。 可以说，深耕海外市场多年，康迪科技对海外用户的需求、市场的规则、服务的痛点都有着深刻认 知。 浙江大学则拥有深厚的科研积淀与顶尖的人才储备。在人工智能、机器人学、计算机视觉等领域，浙 大拥有从理论到应用的全链路创新能力。 双方联手共建联合研发中心，未来也将成功构建"前沿研究-快速孵化-场景验证-迭代优化"的闭环 体系。将实验室里的算法、模型，真正转化为能够在真实泥泞、坡地、暗光环境中稳定运行的机器人 产品。 浙江大学副校长陈刚表示："学校高度重视与康迪科技的合作，此次共建联合研发中心，对进一步贯 彻实施创新驱动发展战略，推动技术成果转化，促进智能安防产业经济社会持续健康发展，都将产生 积极的影响。" 产学研融合新范本。 本文为 ...

Core Viewpoint - LXJ International Holdings Limited, known as "Laoxiangji," is advancing its IPO process on the Hong Kong Stock Exchange, with a focus on expanding its integrated supply chain and store network while enhancing digital capabilities and brand marketing [2][6]. Group 1: Company Overview - Laoxiangji opened its first restaurant in Hefei, Anhui in 2003 and has since developed a "direct + franchise" store network, with 1,658 stores across 61 cities as of August 31, 2025, projected to increase to 1,777 by December 31, 2025 [2]. - The company is recognized as one of the first standardized Chinese fast-food companies in all core operational aspects [2]. Group 2: Market Position - Laoxiangji ranks first in the Chinese Chinese fast-food industry with a market share of 0.9% based on total transaction volume in 2024, and ranks eighth in the overall fast-food industry with a 0.5% market share [2]. - In the East China region, Laoxiangji holds the largest market share of 2.2% in the Chinese Chinese fast-food sector, significantly higher than the second-largest competitor [3]. Group 3: Financial Performance - Revenue figures for Laoxiangji from 2022 to 2024 were 4.528 billion, 5.651 billion, and 6.288 billion yuan, respectively. For the first eight months of 2025, revenue increased by 10.9% to 4.578 billion yuan compared to the same period in 2024 [5]. - Adjusted net profits for the same period were 268 million, 403 million, and 439 million yuan, with a 13.2% increase in the first eight months of 2025 to 394 million yuan [5]. Group 4: Supply Chain and Technology - Laoxiangji is the only major Chinese fast-food company with a fully integrated supply chain, including chicken farming, central kitchens, and restaurant services, and is the first to implement a comprehensive traceability system [4]. - The company has established three chicken farms and two automated central kitchens in Anhui, along with eight distribution centers nationwide to ensure fresh ingredient delivery [4]. Group 5: Membership and Customer Engagement - As of August 31, 2025, Laoxiangji has approximately 31.4 million registered members, with 13 million active members and 600,000 paying members, leading the Chinese Chinese fast-food sector in membership size [5]. - Active members place an average of 8.1 orders, while paying members have a monthly purchase frequency of 5.8 times [5].

Core Viewpoint - Chengdu Zeling Biopharmaceutical Technology Co., Ltd. is preparing for an IPO on the Hong Kong Stock Exchange, focusing on innovative small molecule therapies for unmet medical needs in hematological diseases, oncology, CNS, and immune/inflammatory diseases [3][9]. Product Pipeline - As of January 5, 2026, Zeling Biopharmaceutical has established a pipeline consisting of eight proprietary assets, including two core products, FM and PM, which are in late-stage clinical trials [3][9]. - FM is a novel triple-target inhibitor for treating myeloproliferative neoplasms (MPN), showing a twofold increase in response rate compared to the standard treatment, ruxolitinib [5][6]. - PM is a high-selectivity HDAC inhibitor designed for relapsed/refractory DLBCL, demonstrating impressive clinical efficacy, particularly in subtypes with poor responses to current therapies [7]. Technological Platforms - The company's drug discovery capabilities are driven by three integrated technology platforms: ZeniFold for structural biology, ZeniMind for AI-driven drug discovery, and ZeniScreen for clinical-relevant disease model screening [8]. - These platforms enable a rational, data-driven, and efficient drug discovery process, addressing key limitations of existing therapies [8]. Financial Backing and IPO Plans - Zeling Biopharmaceutical has received investments from notable institutions, completing a C round financing of 511 million RMB in November 2025, with a post-money valuation of 3.411 billion RMB [8]. - The IPO proceeds will primarily fund the R&D of core products FM and PM, other pipeline assets, enhancement of technology platforms, and general corporate purposes [9].

Core Viewpoint - The article discusses the successful development and open-sourcing of the GLM-Image model by Zhiyu and Huawei, marking a significant advancement in training state-of-the-art (SOTA) models on domestic chips, specifically the Ascend Atlas 800T A2 device and MindSpore AI framework [4][10]. Group 1: Model Development and Architecture - GLM-Image is the first SOTA model trained entirely on domestic chips, utilizing a self-regressive structure and a hybrid architecture combining "self-regression + diffusion encoder" to enhance global instruction understanding and local detail depiction [4][5]. - The model addresses challenges in generating knowledge-intensive scenes such as posters and PPTs, representing a step towards a new generation of "knowledge + reasoning" cognitive generative models [5][10]. - The innovative architecture allows GLM-Image to understand complex instructions and accurately render text, integrating a 9B self-regressive model with a 7B DiT diffusion decoder [7]. Group 2: Performance Metrics - GLM-Image achieved top performance in the CVTG-2K benchmark with a Word Accuracy of 0.9116 and a Normalized Edit Distance (NED) of 0.9557, indicating high accuracy in generating text within images [6][8]. - In the LongText-Bench benchmark, GLM-Image scored 0.952 in English and 0.979 in Chinese, leading among open-source models for rendering long and multi-line texts [8]. Group 3: Technical Innovations - The model supports adaptive processing of various resolutions, generating images from 1024x1024 to 2048x2048 without the need for retraining [7]. - GLM-Image's training process was optimized through advanced techniques such as dynamic graph multi-level pipeline dispatch and high-performance fusion operators, enhancing both training stability and performance [10]. Group 4: Market Implications - The introduction of GLM-Image signifies a deep exploration of domestic computing ecosystems, showcasing the potential of domestic computing power in training high-performance multimodal generative models [10]. - The model's open-source nature aims to share technological pathways and practical insights with the open-source community, potentially driving further advancements in the field [5][10].