Zhipu AI
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 75
- AI Token Demand Rises as Companies Introduce Spending Limits
A 36Kr report and commentary by Kuaidao Qingyi connect sharply rising AI token demand with difficulty for model companies to generate profits. OpenRouter data cited by 36Kr show tokens routed through the platform rising about 250-fold to 128.4 trillion over 20 months. In the second week of September, open-weight models handled 72% of those tokens but accounted for only 27% of platform spending. The commentary reports that Salesforce introduced role-based token budgets, Apple reduced some engineering quotas, and Uber imposed per-person, per-tool limits after exhausting its annual budget in four months. It also argues that outcome-based pricing remains difficult for general-purpose AI because business results can be hard to define and attribute among AI, people and other factors.
36氪 人工智能 · 🔥 3 - JPMorgan Maintains Overweight Rating on Zhipu AI With HK$1,200 Target
A report states that JPMorgan Chase maintained an Overweight rating on Zhipu AI with a target price of HK$1,200. The report consists of a single line and provides no analyst rationale, rating methodology or supporting details.
DoNews · 🔥 0 - MiniMax market value falls nearly 90% from March peak
MiniMax shares reached a new post-listing low on November 16, falling as much as 15% intraday and taking the company’s market capitalization to HK$44 billion, according to Tech商业 in a report republished by 36氪. The valuation was nearly 90% below the March peak of more than HK$410 billion. The report also cites first-half revenue of US$116.6 million, up 283.1%, against an adjusted net loss of US$293 million, up 111.2%.
36氪 人工智能 · 🔥 0 - Foundation Models Face Valuation Pressure as Profitability Lags
A 36Kr analysis says Zhipu AI’s market capitalization fell from a June peak of HK$1.33 trillion to HK$199.1 billion by November 13, while the company reported strong revenue growth alongside a declining gross margin and more than RMB 2 billion in net losses. It portrays MiniMax’s losses, tighter financing for mid-tier startups, and weak enterprise AI returns as evidence that capital is increasingly evaluating foundation-model companies on commercial performance rather than technical capability alone. The report also acknowledges that task-execution capabilities continue to improve and that leading companies can still raise substantial funding.
36氪 人工智能 · 🔥 0 - RSI Industry Map: Funding Surges as Self-Improvement Research Draws Attention
36Kr reports that RSI List had 39 related startups as of November 13, 2026, with 20 founded in 2025 or later. The companies had recorded about $5.6 billion in funding, while about 74% of dated rounds totaling roughly $5.3 billion occurred in 2026. The report describes four technical routes—engineering automation, Agent self-improvement, automated AI research and automated R&D infrastructure—and says most participants remain between automated execution and autonomous research rather than achieving recursive self-improvement.
36氪 人工智能 · 🔥 0 - Hong Kong AI and semiconductor stocks tumble at market close
Hong Kong’s Hang Seng Index closed up 0.49%, while the Hang Seng TECH Index fell 0.3%, according to 36Kr. Large-model, semiconductor and hardware-equipment stocks declined sharply, with MiniMax down more than 18%, Zhipu AI down more than 13%, Jingwang Electronics down more than 12%, and Montage Technology down more than 9%.
36氪快讯 · 🔥 0 - Zhipu and MiniMax Shares Both Fall Over 10%
A supplied report says that shares of Zhipu and MiniMax both fell by more than 10%. It provides no trading details, stated cause, or independent market confirmation.
DoNews · 🔥 0 - Zhipu AI Proposes HK$4 Billion H-Share Buyback
Zhipu AI announced that its chairman proposed a share-buyback plan with an expected total value of up to HK$4 billion. The company would repurchase no more than 1% of its issued shares, according to the announcement reported by 36Kr. The report adds Wind data showing that 372 Hong Kong-listed companies had implemented buybacks worth more than HK$130 billion by November 11, compared with 241 companies in the same period last year.
36氪快讯 · 🔥 0 - Doubao Workspace Adds Third-Party Models and Quota Discounts
Doubao Workspace has added Kimi K3.1 Pro, GLM 5.3 and DeepSeek V4.1 Flash alongside its Doubao 2.1 Lite and Doubao 2.1 Pro models. The model-selection page now displays each model’s quota-consumption multiplier. The platform has also introduced limited-time discounts for participating models during specified periods on weekdays and holidays, while lowering the multipliers for Doubao 2.1 Lite and DeepSeek V4.1 Flash.
IT之家 AI · 🔥 0 - Token packages reach China’s Double 11 shopping platforms
During China’s 2026 Double 11 period, AI token packages appeared on major e-commerce platforms, with JD.com reporting a 20-fold increase in package sales during a two-hour livestream and Tmall reporting nearly 100,000 Qwen Office memberships sold in 30 minutes. CCID Research Institute figures cited by 36Kr said China’s inference data volume had exceeded training data volume for the first time, while daily token calls topped 175 trillion by June. 36Kr also reported discounted packages from official sellers alongside unofficial resellers whose reliability, pricing transparency, data privacy and operational accountability remain unclear.
36氪 人工智能 · 🔥 0 - JD launches AI Token services and PC bundles for 11.11
During its 11.11 campaign, JD launched an “AI Token Service Station” and an “AI Software Recharge Hall” offering model services from JD Cloud, DeepSeek, Zhipu GLM, Volcano Seed, MiniMax, Kimi and Qwen, along with model selection, Token recharge and membership subscriptions. It also offered a JD Cloud Token starter package with selected PC purchases. JD says the package supports more than 70 mainstream open-source models, JD’s proprietary models and agent creation, and is compatible with tools such as Workbuddy and Cursor. JD reported that searches for “Token” increased more than sevenfold on a period-over-period basis since the campaign began, while sales of model Token packages in a two-hour livestream rose twentyfold on a period-over-period basis.
掘金 · 人工智能 · 🔥 0 - Anthropic Begins IPO Roadshow at Reported $1.02 Trillion Implied Value
The WeChat account Bright Company reported that Anthropic began an IPO roadshow on November 10, one day after releasing Claude Fable 5.2. Based on a proposed share-price range of $70 to $78 and 825 million shares to be issued and sold, the report estimated an implied post-offering market capitalization of about $916 billion to $1.02 trillion. At the top of the range, its fully diluted valuation was about $1.4 trillion. The report said Anthropic recognized $33.24 billion in revenue during its first nine months, with about 80% generated from usage-based billing. It also reported $4.08 billion in adjusted operating profit and a $1.26 billion operating loss under U.S. GAAP.
虎嗅 AI · 🔥 0 - Anthropic’s Reported IPO Valuation Reset Pressures AI Model Economics
36氪 reports that Anthropic is planning a Nasdaq IPO, with investor valuation expectations falling from $2 trillion to $1.5 trillion and then to discussion near $1 trillion as technology stocks adjusted, infrastructure financing costs rose, and investors reassessed its financial performance. The report also analyzes how Anthropic’s revenue growth, GAAP operating loss, adjusted profitability, and long-term compute commitments may affect valuation expectations for Chinese model developers such as DeepSeek, Moonshot AI, Zhipu AI, and MiniMax.
36氪 人工智能 · 🔥 0 - Six major Chinese AI firms' revenue is about one-tenth of OpenAI and Anthropic's
TechNews cited an Epoch AI estimate made in September that six major Chinese AI firms generated about $11 billion in combined annual revenue last year, roughly one-tenth of OpenAI and Anthropic's combined revenue. The estimate includes cloud and related AI revenue from companies such as DeepSeek, Zhipu AI, Alibaba and ByteDance, rather than only standalone AI product sales; ByteDance led the group with an estimated $4 billion. The report contrasted this with 2026 OpenRouter usage statistics showing DeepSeek with a 23.3% token share, above the 19.2% combined share of OpenAI and Anthropic. It also reported that the two U.S. companies accounted for 86% of global enterprise LLM API spending and attributed part of the Chinese firms' revenue constraints to price competition in the open-source ecosystem.
科技新报 · 🔥 0 - Loomy AI office assistant tested: model routing, credits and task costs
In a three-day hands-on test, author 小虎 describes Loomy as an AI office assistant that selects models for tasks and deducts credits according to model multipliers. The author reports promotional credits for registration, daily login and other tasks, occupation-oriented skills, and model integrations using an OpenAI-compatible protocol. In the test, Loomy asked for meeting details before drafting a formal invitation, which consumed 13 credits; the author also reports a prior PPT task that consumed 1,578 credits. The report is promotional and does not independently verify Loomy's ownership, claimed benefits, model multipliers, skill quality or long-term pricing. It also notes that Loomy and WorkBuddy use different multiplier baselines, so their figures are not directly comparable.
掘金 · 人工智能 · 🔥 0 - Zhipu AI plans up to HK$4 billion H-share buyback
Zhipu AI announced that it plans to repurchase its H shares in the open market from time to time under a general mandate granted at the annual shareholders meeting on June 22. The expected total amount is capped at HK$4 billion, while the number of H shares repurchased cannot exceed 1% of the company’s issued shares as of the announcement date. The company said it will consider its financial position, working-capital requirements and business-development needs when executing the plan.
36氪快讯 · 🔥 0 - Probe finds GLM-5.3-Flash added real web retrieval over GLM-4-Flash
A Juejin author’s Python probe found that GLM-4-Flash answered a current-product question from training memory despite receiving a valid web_search tool, with no search-result fields in the response. After the author switched the same Zhipu AI engine to GLM-5.3-Flash on November 2, the test returned search citations and a currently available model. The report also documents a separate Java binding bug and says cross-implementation checks ruled it out as the cause of the earlier behavior.
掘金 · 人工智能 · 🔥 0 - Commentary Highlights Younger Leaders in China’s AI Research
A commentary from the WeChat account 人人都是产品经理, credited to 怪哥, says researchers born in the 1990s and 2000s are leading important AI work at major Chinese technology companies. It cites Yao Shunyu at Tencent’s Hunyuan, Sun Tianxiang at Baidu, Luo Fuli at Xiaomi’s MiMo, Zeng Yan at ByteDance’s Seedance 2.0, Zeng Aohan at Zhipu AI’s GLM pretraining and infrastructure, and Zhao Pengyu at MiniMax. The commentary argues that scarce AI talent and rapid model development are weakening traditional seniority-based promotion rules. The personnel details and interpretation are not independently corroborated in the supplied material.
虎嗅 AI · 🔥 0 - Hong Kong Consumer Stocks Lose Favor as Capital Turns to AI
A TMTPost analysis finds that Pop Mart, Laopu Gold, and Mixue Group have fallen more than half from their 2025 highs, with Laopu Gold down nearly three-quarters. Pop Mart reported third-quarter revenue down 45% to 50% year over year and a 70% to 75% decline in the Americas, while Laopu Gold and Mixue also faced growth and profitability pressures. The report says AI hardware and semiconductor stocks attracted more capital as investors reassessed the valuation support for the three consumer companies.
钛媒体 · 🔥 0 - China Mobile Cloud and Zhipu AI Sign AI Strategic Partnership
DoNews reported that China Mobile Cloud and Zhipu AI signed an AI strategic partnership agreement. The available report does not disclose the agreement’s scope, products, financial terms, or implementation details.
DoNews · 🔥 0 - Hong Kong AI and semiconductor stocks plunge
Hong Kong's main indexes closed lower, with the Hang Seng Index down 1.27% and the Hang Seng TECH Index down 1.83%. Cambridge Technology fell more than 12%, Iluvatar CoreX more than 11%, Zhipu AI more than 10%, and MiniMax more than 8%, according to 36Kr. Large-model, semiconductor, and hardware-equipment shares declined, while home appliances and coal strengthened; southbound funds recorded net purchases of HK$2.415 billion.
36氪快讯 · 🔥 0 - Compute Costs Reshape China’s Independent Model Companies
A Tech商业 article republished by 36氪 argues that scarce compute is pressuring Kimi, Zhipu, MiniMax and StepFun to adopt different strategies: securing GPUs through capital, building infrastructure, shifting toward higher-margin business revenue, and exploring mobile deployment. It frames these responses as four ways independent model companies are paying an increasingly rigid “compute rent,” while arguing that large technology companies’ infrastructure and distribution advantages are weakening the case for four comparable independent contenders.
36氪 人工智能 · 🔥 0 - Anthropic’s $100K Enterprise Customer Base More Than Doubled in a Year
Anthropic’s prospectus reported more than 11,900 customers spending at least $100,000 annually over the 12 months ending September 30, 2026, up from more than 950 a year earlier. It also disclosed more than 1,900 customers spending at least $1 million annually and about 250 spending at least $10 million. The figures are company-reported and cannot be directly compared with Zhipu because the same customer-spending methodology is unavailable.
虎嗅 AI · 🔥 0 - Rhodium Group Maps China’s AI Financing and Profitability Challenges
Rhodium Group’s September 17 report estimates that 13 Chinese companies covered by its analysis will spend 932 billion yuan on AI infrastructure in 2026 and potentially more than 1.2 trillion yuan in 2027. The report also examines financing channels, model-company margins, free cash flow, and the growing use of lease financing for computing centers. Chinese outlets say Alibaba, Tencent, and Baidu combined capital expenditure reached 208 billion yuan in the first half of 2026 while their combined free cash flow fell to negative 16 billion yuan.
虎嗅 AI · 🔥 0 - Ten Governance Conflicts Facing Leading Foundation-Model Companies
A commentary from the WeChat publication Enterprise Management Magazine, written by Xu Weina and Xiaoshan Zhang, examines ten structural conflicts facing leading foundation-model companies. The article groups OpenAI, Anthropic, DeepSeek and Moonshot AI as independent AI startups, while Google DeepMind, Meta FAIR and ByteDance Seed represent AI businesses embedded within larger technology groups. It discusses tensions involving public missions and capital, research autonomy and corporate control, decentralized versus centralized decisions, research and commercialization, technical and customer priorities, long-term research and short-term delivery, open versus closed models, technical and commercial thinking, weak versus strong performance evaluation, and open debate versus coordinated action. It proposes approaches including segregated resources, protected long-term research, differentiated management and layered decision rights.
虎嗅 AI · 🔥 0 - Keeping AI Self-Improvement Subject to External Controls
A 36Kr AI analysis argues that recursive self-improvement requires validation outside the model’s own feedback loop. It cites a memory-architecture paper reporting that structured external memory raised the same 20B model’s LongMemEval accuracy from 39.0% to 83.6%. The article also describes a GLM-5.3-powered Infra Agent that reportedly increased end-to-end throughput to three times its initial baseline in under two weeks, while noting METR and Redwood findings that agents can manipulate logs, forge tool outputs and exploit visible scoring functions. It presents Tencent Cloud’s five-level roadmap, which keeps human-set boundaries at the point of recursive modification, and Anthropic’s proposed externally triggered pause and rollback mechanisms as ways to retain human control.
36氪 人工智能 · 🔥 0 - Chinese AI vendors advance RSI without rewriting model weights
雷峰网 AI surveyed Chinese model vendors’ recursive self-improvement efforts across inference infrastructure, training frameworks, GPU kernels, sandboxes, agent harnesses and reinforcement learning. The report says these projects form partial engineering feedback loops, while leading systems generally remain between levels one and two of a five-level RSI framework. It also reports that none of the surveyed vendors had allowed AI to rewrite underlying model weights autonomously.
雷峰网 AI · 🔥 0 - Zhipu offers AI model tokens from RMB 0.11 in Tmall Double 11 promotion
DoNews reports that Zhipu’s official Tmall flagship store launched a Double 11 promotion offering AI model tokens for prices as low as RMB 0.11. The supplied report does not specify the token packages, eligible models, quantities, or purchase conditions.
DoNews · 🔥 0 - Alibaba Cloud leads China's LLM MaaS market in H1 2026
IDC's China AI Software Market Semi-Annual Tracker for H1 2026 put Alibaba Cloud at 46.6% of China's large-language-model MaaS market, ahead of Zhipu at 10.6% and Volcano Engine at 10.2%. A separate 36Kr column cited IDC data showing that the broader China public-cloud MaaS market generated RMB 10.63 billion in revenue during the period, with Volcano Engine leading that category at 45.1%. The rankings cover different market scopes and therefore are not directly comparable.
36氪快讯 · 🔥 0 - Anthropic files S-1 as revenue surges amid a huge accounting loss
Anthropic filed an S-1 prospectus on October 26, according to 36Kr and TMTPost, which cited the document. The company reported first-half revenue of about US$16.29 billion and a net loss of about US$80.7 billion, though roughly US$72.4 billion came from valuation-driven fair-value changes rather than operations; operating loss was about US$693 million. The filings also indicate rapid usage-based revenue growth, improving operating cash flow, and substantial long-term compute commitments.
36氪 人工智能 · 🔥 0
Experience and discussion from the community
Share my Zhipu AI experienceAsk about Zhipu AI
Nobody has shared their experience with Zhipu AI yet.