智谱万亿了,DeepSeek还会远吗?

marsbit2026-06-22 tarihinde yayınlandı2026-06-22 tarihinde güncellendi

Özet

近日,中国大模型公司智谱在港股盘中市值突破1万亿港元,年内涨幅超1900%,总市值达1.1万亿港元,接近阿里巴巴一半、美团两倍。上市不到半年,其市值从500亿港元冲入万亿区间,引发市场对中国AI估值边界的广泛关注。 智谱市值飙升的直接催化剂是其GLM-5.2模型,市场将其视为“国产Anthropic”。全球AI估值锚因Anthropic在代码与企业Agent场景的商业化成功而抬升,智谱借此被重估。同时,马斯克与智谱创始人唐杰的隔空互动,进一步强化了国产模型参与全球前沿竞争的形象。 除智谱外,中国大模型行业资本动作频繁:DeepSeek新一轮融资后估值超500亿美元;MiniMax启动A股IPO;阶跃星辰和月之暗面也瞄准港股。港股市场正成为硬科技资产的再定价场所,资金集中涌入AI等稀缺赛道。 然而,智谱2025年营收仅7.24亿元,净亏损达47.18亿元,万亿市值引发“神话还是泡沫”的争论。市场一边承认估值高昂,一边追逐头部标的,泡沫与溢价共存。7月上旬,智谱和MiniMax将迎来限售股解禁,可能加剧股价波动。 行业的核心挑战在于商业化能力。智谱2025年收入同比增长131.9%,主要来自MaaS平台与企业级业务,并于2026年一季度对API服务提价。但研发开支高达31.8亿元,为收入的4倍多,亏损压力显著。整个中国市场调用量快速增长,IDC预计2026年MaaS调用量将达40000万亿Tokens,C端AI应用月活用户已超4.4亿,但用户规模不等同于盈利。 Anthropic预计在2026年二季度实现经营利润,若成真,将成为行业商业化转折的关键参照。中国大模型公司需证明能将模型能力转化为持续收入与现金流,才能支撑当前打开的估值边界。 智谱突破万亿市值是中国AI产业重估的标志性高点,但能否守住并形成行业曲线,最终取决于国产大模型厂商的商业化成果。

近日,一家中国基模公司刷新了市值纪录。

智谱盘中市值突破1万亿港元,年内涨幅超过1900%。截至收盘,这家公司总市值来到了1.1万亿港元,约相当于阿里巴巴一半,接近美团两倍、京东三倍。

上市不到半年,智谱从500亿港元级别公司,冲进港股万亿市值区间。

在港股“万亿俱乐部”门槛之上,是像腾讯、阿里、中国移动、工商银行这样的产业巨头,背后都有经过验证的用户入口、现金流或基础设施属性。

智谱进入这个坐标系,意味着公开市场对中国大模型公司的定价正在不断膨胀。

不过,过山车一般的增长速度,背后是一场神话还是泡沫的争论。

智谱上市的第一份财报,披露了2025年全年的营收数据:收入7.24亿元,净亏损47.18亿元,研发开支31.8亿元。

一家上一年收入7亿元的公司,市值却先到了万亿级。看多者把它视作中国AI产业的估值突破;而看空者认为,当前股价的泡沫属性较强,已经跑在基本面前面太远。

某种程度上来说,这场争论已经超出智谱,而是上升到中国大模型产业估值本身。

最新消息显示,DeepSeek在新一轮融资后估值超过500亿美元,而阶跃星辰和月之暗面的资本化预期不断升温,双双瞄准港股市场。

中国大模型公司的估值曲线,正在从一级市场抬到二级市场。伴随着智谱成为第一家万亿市值大模型企业,这条曲线正在不断向上突破新界限。

1

智谱这一轮市值攀升,直接催化剂是GLM-5.2。

去年以来,行业内的主流模型升级,关注点都集中在Coding、Agent、长程任务和复杂工程能力。

在这些场景下,智谱GLM系列可以跻身第一梯队,而全球最强参照是Anthropic。

今年以来,Anthropic已完成650亿美元融资,投后估值9650亿美元。它被追捧的核心,主要来自Claude Code在开发者场景里跑出收入。

官方披露数据显示,Claude Code年化收入已经超过25亿美元,较年初翻倍。

这个样本给全球AI公司更换了估值锚。

C端聊天产品证明用户规模,代码和企业Agent证明付费能力。前者制造声量,后者靠近收入;而收入背后的商业化能力,正是资本市场评估大模型企业的关键。

智谱这一轮被重估,正是因为市场把它放进了“国产Anthropic”的坐标里观察。

近日,特斯拉首席执行官埃隆·马斯克和智谱创始人唐杰的隔空互动,放大了这层想象。

马斯克判断,中国可能在2027年一季度出现接近Anthropic Fable级别的模型。

唐杰回应称,不会等那么久。尽管这个回复本身简短得没什么信息,但是侧面凸显出,国产模型已经身处全球前沿模型追赶叙事里,就看谁能跟上头部竞争对手的脚步。

今年以来,除了智谱的股价攀升,国产模型圈资本动作不断。

智谱在本月初通过A股发行议案,拟登陆科创板募资不超过150亿元,其中120亿元投向通用基座大模型研发,20亿元投向MaaS一站式服务平台。

几乎同一时间,MiniMax与中信证券签署辅导协议,正式启动A股IPO进程,并评估发行人民币股份登陆科创板。两家港股大模型公司同时推进A股路径,说明中国AI估值曲线正在从港股弹性,延伸到A股产业融资。

与此同时,DeepSeek在一级市场疯狂吸纳融资。

近期,DeepSeek完成超过74亿美元融资,估值超过500亿美元。这家来自杭州的AI巨头,依靠出色的模型效率、全球开发者声量和开源生态,把中国大模型一级市场价格推到新高度。

而在二级市场,在智谱和MiniMax相继登陆后,港股市场也在转向拥抱AI产业。

今年以来,港股已有65家公司上市,合计募资超过1765亿港元。硬科技和高端制造企业贡献近1293亿港元,占比超过七成。半导体、电气设备、AI公司成为发行主线,资金正在把港股当作中国硬科技资产的再定价市场。

换言之,智谱的万亿市值,不只来自GLM-5.2的出色表现,也来自几个条件同时出现:全球AI资产估值上扬,中国大模型公司资本化提速,港股缺少纯正大模型标的,硬科技资金重新聚集。

所以,市场真正买入的,是中国AI估值边界仍能继续拓宽。只是,这条边界拓得越宽,关于产业泡沫的争论就会持续下去。

2

看到智谱的万亿市值,大部分投资者直觉上判断,这里面必然有泡沫的成分在。

一家年收入7亿多元的公司,盘中市值超过万亿港元,任何估值模型都会感到紧张。

更何况,智谱股价此前已出现大幅波动。5月末,它盘中市值曾突破8800亿港元,随后快速回落;今日再度冲上万亿的过程中,盘中震荡依旧剧烈,短短十几分钟时间从高点回落,蒸发了上千亿市值。

不过,这也正是AI行情最特殊的地方。市场一边承认价格昂贵,一边继续追逐头部标的。

泡沫警告和估值上调同时发生,构成这一轮AI资本周期的底色。

与此同时,港股新股市场已经出现分化。

今年一季度,香港IPO破发率升至44.74%,明显高于去年。传统行业和认购不足的公司承压,硬科技、AI、新能源企业仍被资金追逐。资金没有无差别买入新股,它只在少数赛道集中冒险。

显然,以智谱为代表的AI产业,属于被集中冒险的资产。突破万亿,靠的是市场对中国AI基础设施公司的稀缺预期。

只是,资本不会无止境投入。7月上旬,智谱和MiniMax将迎来上市后首轮限售股解禁。

智谱的解禁时间是7月8日,规模约2568.16万股,占总股本约5.76%,按近期股价测算,对应解禁市值约269亿港元。解禁前,智谱流通股约1735.08万股。换句话说,即便总股本占比不高,新增可交易筹码也会明显改变此前低流通盘结构。

MiniMax压力更大。它将在7月9日解禁约1.07亿股,占总股本约34.25%。其中,财务型投资者持有比例超过三分之一。

两家公司的股价较上市时都经历了大幅上涨,市场自然会提前担心减持压力。

事实上,不只是国产大模型,全球市场也在经历同样的估值冲突。

马斯克和唐杰隔空对话前两周,SpaceX上市后市值迅速突破2万亿美元,股价也攀升至210美元区间。

这家航天和卫星互联网公司,上市前夕就大量整合资源,强化其在AI基础设施、xAI协同、星链网络和未来算力的叙事中的行业地位。

针对这家公司,纽约大学斯特恩商学院教授达莫达兰给出的估值低于市场定价,并提醒SpaceX高估值已包含大量未来假设。

而在过去几个交易日中,SpaceX股价经历了大幅回调,股价来到了170美元区间。

6月初,Scion Asset Management创始人、电影《大空头》原型投资人迈克尔·伯里(Michael Burry)也把质疑指向Anthropic。

他认为,Claude背后的大模型研发成本过高,近万亿美元估值很难只靠当前商业模型支撑。

行业内如此大的认知分歧,其实看似对立的观点并不矛盾,在产业高速增长周期,泡沫已经天然成为AI估值的一部分。

决定这部分“泡沫”是否是长期溢价的关键因素,就取决于AI企业的商业化能力。

3

突破万亿市值后,智谱眼前真正的问题,是收入的增长速度。

2025年,智谱收入7.24亿元,同比增长131.9%。其中,本地化部署收入5.34亿元,同比增长102.3%;云端部署收入1.90亿元,同比增长292.6%;企业级智能体业务收入1.66亿元,同比增长248.8%。

其中,智谱MaaS API平台ARR约17亿元,过去12个月提升60倍,平台毛利率提升至18.9%。这意味着MaaS的收入占比正在不断增加。

2026年一季度,面对模型调用的旺盛需求,智谱直接选择了——涨价。

今年2月以来,智谱对GLM Coding Plan套餐价格进行结构性调整,整体涨幅30%起,理由是用户规模和调用量快速提升,需要保障高负载下的稳定性和服务质量。

只是,这部分营收究竟对应着多少收入增长,就要到智谱的下一份财报中揭晓了。

与此同时,作为一家独立模型公司,智谱的成本压力巨大。

2025年,智谱净亏损47.18亿元,经调整净亏损31.82亿元,研发开支31.8亿元。研发投入相当于收入的4倍多。

模型公司可以烧钱,但是必须让资本市场看到长期盈利的预期。

好在,整个中国市场的调用量基础正在不断抬升。

IDC数据显示,2024年中国企业级MaaS市场调用量为114万亿Tokens,2025年跃升至1944万亿Tokens,预计2026年达到40000万亿Tokens。Token正在从技术指标变成商业结算单位。

国产模型体量最头部的字节,旗下豆包大模型日均Token调用量已突破120万亿,三个月翻倍,较2024年5月发布时增长1000倍。累计调用量超过一万亿Token的企业,从去年底100家增至140家。

与此同时,C端场景也在扩容。

QuestMobile数据显示,2026年3月,中国AI原生App月活用户规模达到4.4亿。豆包、千问、DeepSeek位列前三。AI应用已经从小圈层工具变成大众产品。

只是,用户规模并不代表利润水平,行业内的泡沫担忧,本质上是对AI无底线烧钱、看不到盈利预期的无力感。在这个问题上,Anthropic仍然是那个最强参照。

今年二季度,Anthropic预计收入109亿美元,并实现约5.59亿美元经营利润。如果这一预期兑现,它将成为全球AI公司商业化转折样本。市场会因此相信,大模型公司不一定长期停留在“烧钱换规模”的阶段。

这也是中国AI公司被重估的原因。

如果Anthropic能靠代码、Agent和企业API接近盈利,国内市场会自然寻找中国对应标的。智谱、DeepSeek、MiniMax、月之暗面,以及几家头部AI大厂,都将被放进这条商业化曲线里比较。

随着智谱率先突破万亿,这也成为了中国AI产业接下来的共同命题。

估值边界已经打开,关于泡沫的争议不会停息,市场的波动和不确定性依然存在。

要巩固这条估值边界,大模型厂商首先要把模型能力稳定转成收入,把收入继续推向毛利和现金流;同时,也需要更多国产厂商在追赶Anthropic的路上跑出有效样本。

智谱今天站上万亿市值,是中国AI产业重估的第一个高点。这个高点能不能守住,能不能变成一条行业曲线,最终要由国产大模型厂商的商业化成果来回答。

本文来自微信公众号“字母榜”,作者:李炤锋

Trend Kriptolar

İlgili Sorular

Q智谱市值突破万亿港元,其2025年全年的营收和净亏损分别是多少?

A根据文章,智谱2025年全年收入为7.24亿元,净亏损为47.18亿元。

Q文章中提到,智谱市值冲上万亿港元的直接催化剂是什么?

A智谱这一轮市值攀升的直接催化剂是其模型升级,具体是GLM-5.2的发布和出色表现。

Q文章列举了哪些因素共同促成了智谱的万亿市值?

A智谱的万亿市值是几个条件共同作用的结果:全球AI资产估值上扬、中国大模型公司资本化提速、港股市场缺少纯正的大模型投资标的、以及硬科技资金在港股市场的重新聚集。

Q根据文章,中国AI原生App的月活用户规模在2026年3月达到了多少?

A根据QuestMobile数据,2026年3月,中国AI原生App月活用户规模达到了4.4亿。

Q文章指出,要巩固已经打开的AI产业估值边界,大模型厂商需要做哪两件核心事情?

A要巩固这条估值边界,大模型厂商首先要把模型能力稳定地转化为收入,并进一步将收入转化为毛利和现金流;同时,也需要更多的国产厂商在追赶行业标杆(如Anthropic)的过程中,跑出有效的商业化样本。

İlgili Okumalar

OpenAI Partners with PE Firms, Investing $4 Billion. Let's Talk About Silicon Valley's Hottest New Role: FDE.

The hottest new role in Silicon Valley is the Forward Deployment Engineer (FDE), a hybrid of engineer and business consultant whose core mission is to transform AI demos into native, practical workflows within client organizations. The recent surge in demand is driven by a strategic shift from leading AI companies. OpenAI, partnering with 19 private equity firms in a $4 billion investment, formed a Deployment Company and acquired Tomoro along with its 150 FDEs. Anthropic also announced a $1.5 billion joint venture with financial institutions like Blackstone. The article, based on interviews with industry experts Jove (FDE lead at Cresta) and Oliver (VP at Invisible Technologies, ex-McKinsey), explores the FDE role and the rise of deployment-focused companies. Key insights include: **The FDE Role:** Jove describes an FDE as a "Forward Deployed CTO"—a technically strong engineer who works intimately with clients to implement AI solutions, learn from the process, and feed those insights back to improve the core product. They require expertise in AI agents, client-facing experience, resilience, and the ability to handle complex, imperfect systems. While AI tools enhance their efficiency, the role's complexity makes full automation a distant prospect. **Industry Shift:** Model companies are moving beyond selling tools to ensuring real-world adoption. This blurs the line between model and application companies. Collaborations with private equity (PE) firms are key, providing access to large portfolios of traditional businesses needing AI transformation. For PE firms, these partnerships offer signal value to LPs, create tangible value in portfolio companies, and provide exposure to high-growth AI assets. **Consulting & Transformation:** AI deployment involves deep, customized workflow redesign, moving beyond simple tool augmentation. Companies like Invisible Technologies build modular platforms to create bespoke, AI-native workflows for clients. While traditional consulting will see growth in helping businesses rethink their models for AI, the real value is captured by firms that leave behind transformed, operational systems. Critical success factors include building robust data foundations and strategically deciding which workflow steps should be deterministic versus AI-driven. The ultimate goal shifts from pure cost-cutting to unlocking new revenue opportunities previously impossible without AI-scale capabilities.

marsbit6 dk önce

OpenAI Partners with PE Firms, Investing $4 Billion. Let's Talk About Silicon Valley's Hottest New Role: FDE.

marsbit6 dk önce

Why Is DeFi Insurance Unpopular?

The article explores the core reasons why DeFi insurance remains largely unutilized despite its potential to eliminate traditional insurance inefficiencies and malicious claim denials through automated smart contracts. Key points include: 1. **Low Adoption & Minimal Payouts:** Leading provider Nexus Mutual has paid only ~$18M in claims since 2019, dwarfed by single hack losses (e.g., Kelp DAO's $292M loss). 2. **High Correlation Risk:** Unlike traditional insurance (e.g., house fires), DeFi risks (oracle failures, bridge hacks) are systemic and can simultaneously impact multiple protocols, threatening to drain entire insurance pools. 3. **Prohibitive Cost vs. Reward:** For many protocols (Aave, Morpho, Compound), insurance premiums (1.5%-6%) consume a significant portion or even all of the native yield (3%-4%), leaving investors with meager returns. In some cases (Maple Finance, Ethena), premiums can even result in net-negative yields. 4. **Inadequate Capacity:** The total DeFi insurance pool (e.g., Nexus Mutual's $81.56M) is minuscule compared to the hundreds of billions in total value locked (TVL), creating a massive supply-demand gap. 5. **Structural Flaws:** The claims assessment model (e.g., Nexus Mutual's member voting) creates a conflict of interest, as voters bear the loss if a claim is paid. There is also no regulatory mandate forcing DeFi protocols to obtain insurance. The industry is adapting by focusing on preventative measures (e.g., bug bounty coverage) and seeking external capital via reinsurance. However, the fundamental issues of small pool size, correlated risk, and misaligned economic incentives persist. The article concludes that DeFi insurance, like a public lighthouse, provides shared security benefits, but if everyone relies on others to pay for it, no one will, leaving the ecosystem vulnerable.

Foresight News20 dk önce

Why Is DeFi Insurance Unpopular?

Foresight News20 dk önce

Ethlabs Founded, Treasury Companies to Fund Ethereum Post-EF

Former Ethereum Foundation (EF) core researchers Ansgar Dietrichs, Barnabé Monnot, Caspar Schwarz-Schilling, Josh Rudolf, and Julian Ma announced the launch of Ethlabs, an independent non-profit R&D lab focused on Ethereum core protocol research and institutional-grade infrastructure. The initiative, backed by over 50 community participants including ETH treasury companies BitMine and Sharplink, Joseph Lubin, Hayden Adams, and Jesse Pollak, aims to make Ethereum the global economic settlement layer. This move comes amidst significant pressure on the EF, which has seen key departures and a strategic narrowing of its focus. A critical funding gap of approximately $30 million annually for core client development, following the expiration of the client incentive program, poses a near-term risk to the network's development. The context includes the evolution of ETH's value narrative. While mechanisms like EIP-1559 and the Merge previously supported the "ultrasound money" thesis, the success of L2 scaling via EIP-4844 has drastically reduced L1 fee revenue, leading to net ETH issuance and challenging that narrative. Ethlabs has listed ETH monetary economics as a primary research focus. Backing from corporate ETH treasuries like BitMine and Sharplink represents a strategic alignment, as these entities' asset values are directly tied to Ethereum's health and adoption. Their support is an investment rather than a pure donation. Ethereum's governance is shifting from a centralized EF model to a distributed network of specialized "manager nodes," including Ethlabs and a streamlined EF. While this promotes efficiency and reduces single-point failure risk, it introduces new challenges in coordination, priority alignment, and filling critical funding gaps across the decentralized ecosystem.

Foresight News44 dk önce

Ethlabs Founded, Treasury Companies to Fund Ethereum Post-EF

Foresight News44 dk önce

İşlemler

Spot
Futures

Popüler Makaleler

GLM Nasıl Satın Alınır

HTX.com’a hoş geldiniz! Golem (GLM) satın alma işlemlerini basit ve kullanışlı bir hâle getirdik. Adım adım açıkladığımız rehberimizi takip ederek kripto yolculuğunuza başlayın. 1. Adım: HTX Hesabınızı OluşturunHTX'te ücretsiz bir hesap açmak için e-posta adresinizi veya telefon numaranızı kullanın. Sorunsuzca kaydolun ve tüm özelliklerin kilidini açın. Hesabımı Aç2. Adım: Kripto Satın Al Bölümüne Gidin ve Ödeme Yönteminizi SeçinKredi/Banka Kartı: Visa veya Mastercard'ınızı kullanarak anında Golem (GLM) satın alın.Bakiye: Sorunsuz bir şekilde işlem yapmak için HTX hesap bakiyenizdeki fonları kullanın.Üçüncü Taraflar: Kullanımı kolaylaştırmak için Google Pay ve Apple Pay gibi popüler ödeme yöntemlerini ekledik.P2P: HTX'teki diğer kullanıcılarla doğrudan işlem yapın.Borsa Dışı (OTC): Yatırımcılar için kişiye özel hizmetler ve rekabetçi döviz kurları sunuyoruz.3. Adım: Golem (GLM) Varlıklarınızı SaklayınGolem (GLM) satın aldıktan sonra HTX hesabınızda saklayın. Alternatif olarak, blok zinciri transferi yoluyla başka bir yere gönderebilir veya diğer kripto para birimlerini takas etmek için kullanabilirsiniz.4. Adım: Golem (GLM) Varlıklarınızla İşlem YapınHTX'in spot piyasasında Golem (GLM) ile kolayca işlemler yapın.Hesabınıza erişin, işlem çiftinizi seçin, işlemlerinizi gerçekleştirin ve gerçek zamanlı olarak izleyin. Hem yeni başlayanlar hem de deneyimli yatırımcılar için kullanıcı dostu bir deneyim sunuyoruz.

101 Toplam GörüntülenmeYayınlanma 2024.12.10Güncellenme 2026.06.02

GLM Nasıl Satın Alınır

Tartışmalar

HTX Topluluğuna hoş geldiniz. Burada, en son platform gelişmeleri hakkında bilgi sahibi olabilir ve profesyonel piyasa görüşlerine erişebilirsiniz. Kullanıcıların GLM (GLM) fiyatı hakkındaki görüşleri aşağıda sunulmaktadır.

活动图片