op

According to Cointelegraph, former Ripple Chief Technology Officer David Schwartz stated that he has sold the vast majority of his cryptocurrency holdings due to a lower personal risk tolerance.Schwartz revealed that his cryptocurrency asset holdings have decreased from approximately 26 million XRP, 1,500 BTC, and 40,000 ETH to about 3 million XRP, 1 BTC, and 6 ETH currently.

X platform U.S. stock trader 3X Long Labubu (@labubu_trader) posted that they bought shares of Anthropic at a valuation of $18 billion during the liquidation sale of FTX's bankruptcy, jokingly thanking "the greatest angel investor" SBF.Previously, according to FT reports, underwriters for Anthropic were preparing to raise the IPO price to above the $70-$78 range, potentially increasing the valuation to over $1.5 trillion. Based on this calculation, the valuation would rise approximately 83 times from the $18 billion purchase valuation.Public information shows that 3X Long Labubu is an AI engineer in Silicon Valley, engaged in AI reasoning and training-related work, and trades AI and semiconductor U.S. stocks as a hobby, with about 77,000 followers on X. SBF's FTX and Alameda invested approximately $500 million in Anthropic in 2021, and after FTX's bankruptcy, its estate management sold this portion of shares in batches in 2024.

The locally developed AI Agent Alma, led by independent developer Yetone, ranked third in the latest round of cross-evaluation initiated by the AI Agent infrastructure company Runta, both in task success rate and cost of success, outperforming tools like Claude Code and OpenCode. The current leaderboard snapshot shows that Alma has a task success rate of 72.7%, just behind Codex CLI at 77.9% and Pi at 72.8%, while Claude Code stands at 70.0%.Alma costs approximately $0.80 on average for each successfully completed task, compared to Claude Code's $0.94, making it about 16% cheaper. This FrontierHarness v1.1 comparison evaluates the performance of the same large model under different Agent execution frameworks, covering 21 sets of Harness configurations, 5 large models, and 62 tasks, with each task repeated 4 times, planning to run a total of 26,040 times, which is 72 times the scale of the previous version. All tools operate in a unified cloud environment, comparing task success rates, token costs, and execution speeds.Alma emphasizes local operation and long-term memory, capable of reading and modifying code, executing terminal commands, searching the web, and supporting Skills and MCP, while also allowing for switching underlying models. Recently, Alma integrated with Herdr, allowing programming Agents like Claude Code, Codex, and Pi to continue operating within Herdr, and enabling users to view execution processes, respond to questions, and approve operations in a centralized chat interface. Users can also remotely manage multiple Agents on their computers via iPhone. The evaluation is still ongoing, and rankings may change; the models being tested are limited to open-source models like GLM, DeepSeek, Kimi, and Muse, and do not cover closed-source models from the GPT and Claude series.

According to the official announcement, the Bitget IPO Prime Anthropic VIP round subscription will close today at 16:00 (UTC+8). VIP 1 - VIP 7 users can participate in the subscription using USDT or USDGO on the IPO Prime page before the deadline.The VIP round subscription period is from November 5 at 16:00 to November 11 at 16:00 (UTC+8), with an investment range of $100 to $3,000,000. The total cap for the investment pool is $1 billion, with a historical reference allocation rate of 40%-80%. The underwriting service fee is 12%, and the distribution time is 3 to 5 working days after the IPO.In addition, the public round subscription opened on November 10 at 16:00 and will continue until 16:00 (UTC+8) on the day before the IPO, open to all users, with an investment range of $100 to $25,000. Eligible users can participate in both rounds simultaneously, with independent allocations for each round.It is reported that Anthropic launched its IPO roadshow last night, planning to issue 825 million shares at a price of $70 to $78 per share. The prospectus reveals that seven companies, including Google and Nvidia, plan to participate in a total of $26 billion in concurrent private placements at the final IPO issuance price.

Financial infrastructure service provider OpenPayd announced that its unaudited revenue for the first half of fiscal year 2027 (from May to October 2026) was 47.5 million USD, with revenue in the second fiscal quarter increasing by 44% year-on-year to 25.4 million USD. In the first half of the fiscal year, the fiat currency infrastructure business contributed 70% of the year-on-year revenue growth from banking and payment services, while the remaining 30% came from stablecoin business (excluding interest income), with stablecoin revenue increasing to 3.9 million USD.OpenPayd also disclosed that its annualized revenue, calculated by multiplying the revenue for the month of October by 12, reached 107 million USD. Previous reports indicated that OpenPayd plans to go public on NASDAQ by the end of the year, with a valuation of up to 1.1 billion USD.

According to the latest stock spot exchange rankings released by RootData, the top three exchanges are Binance, Bitget, and Gate.Among them, Binance scored 82, with a 24h public order book transaction volume of 334 million USD and a market share of 24.24%; Bitget scored 71.6, with a 24h public order book transaction volume of 83.3861 million USD and a market share of 6.04%; Gate scored 70.1, with a 24h public order book transaction volume of 65.8801 million USD and a market share of 4.77%.

According to on-chain analyst Yu Jin's monitoring, the whale @fiyopi recently opened a long position of 24,500 ETH (worth 50.59 million USD) on Hyperliquid at 2,207.5 USD, currently facing a loss of 3.5 million USD. One of their addresses is named "Machi is ugly dog."

The Nikkei 225 index rose for the third consecutive trading day, as related data and corporate earnings reignited optimism in the artificial intelligence sector. A Reuters survey showed that Japan's manufacturing confidence index rose to its highest level in over five years in November.Despite U.S. stocks closing lower, AI cloud computing company CoreWeave expects to achieve profitability in the first half of 2028. Nomura Securities strategist Akihiro Akiyama stated that investors are focused on Nvidia's earnings report next week, and if it performs strongly, it may drive up the stock prices of Japanese AI-related companies.

Alibaba announced that the new generation of the Qwen large model, Qwen4, has begun open testing, with the flagship preview version Qwen4-Max launched. This model has been integrated with Alibaba's Token Plan, Qwen Office, Qoder, Qwen App, and QwenStudio.This generation focuses on upgrading complex programming, front-end design, and agent capabilities. Alibaba also emphasizes professional office and software operations, hoping the model can autonomously use tools and complete multi-step tasks.Qwen4 has been in a warm-up phase for several months. In September of this year, Alibaba announced that the Qwen4 based on the new architecture is in training. The earlier released Qwen3.8-Flash-Next has already opened an experimental version of the new architecture. According to the previously showcased product plan, the Qwen4 series also includes Flash, Plus, and a 27B version aimed at local deployment.

According to a report by the Science and Technology Innovation Board Daily, there is a significant price disparity in the embodied intelligence data collection industry. For the same 1-hour period, the listed price for high-quality real machine data can reach 300 to 1000 yuan, while the hourly wage for end collection workers is as low as 8 yuan, with some original data transactions even dropping to 3 to 5 yuan per hour.The investigation found that the price difference mainly stems from multi-level subcontracting and the resale of data copies. A hundred-yuan order from the client, after being passed through five or six intermediaries, leaves the grassroots collectors with only about 15 yuan. The baseline remuneration for county-level projects is around 30 yuan/hour; falling below this standard poses compliance or source risks. Additionally, the resale of non-original data three to five times is "very common" in the industry, with some teams monetizing the same material through private copies or additional equipment for simultaneous recording.The report points out that while the industry loudly claims that high-quality data is scarce, data collection service providers are valued at tens of billions; on the other hand, low-priced copies are rampant, and data traceability has failed. Some intermediaries gain pricing power through order control, with domestic data producers averaging a gross profit of about 20% to 30%.

Anthropic has publicly submitted and updated its prospectus (S-1/A) to the U.S. Securities and Exchange Commission, planning to go public on Nasdaq under the stock ticker ANT, with Goldman Sachs and Morgan Stanley serving as lead underwriters. The latest amendment submitted on November 10 reveals an offering price range of $70 to $78 per share, with approximately 825 million shares of Class A common stock being offered, of which the company will issue about 801.35 million shares, existing shareholders will sell about 23.65 million shares, and underwriters will have an additional overallotment option of about 123.75 million shares. The company will also allocate approximately $26 billion of non-voting Class N shares to strategic investors such as Google, Nvidia, AMD, Broadcom, Micron, Microsoft, and TSMC, priced in line with the IPO price.The market expects pricing around November 18, with trading potentially starting on November 19. Based on approximately 1.309 billion shares post-offering, this corresponds to a valuation of about $917 billion to $1.02 trillion; accounting for options, restricted stock, and other fully diluted factors, the high-end valuation could reach about $1.4 trillion. Previously, the market anticipated a fundraising scale of about $100 billion, with a maximum valuation reaching $2 trillion, but the latest terms have narrowed.FTX-affiliated Alameda Research acquired approximately 7.84% of Anthropic for about $500 million between 2021 and 2022, corresponding to about 44.54 million shares of Series B. After FTX filed for bankruptcy protection in November 2022, the bankruptcy estate sold about 29.5 million shares in March 2024 for a total transaction amount of about $884 million, and in June sold another approximately 15 million shares, recovering about $452 million, totaling about $1.34 billion. According to the latest prospectus's fully diluted share count, the original ownership percentage has decreased to about 2.5%, corresponding to a theoretical value of about $3.11 billion to $3.47 billion based on the IPO price range.

According to a report by Reuters, three informed sources stated that Microsoft and other companies and industry groups are urging the Trump administration to stop plans to ban the import of key components for AI data centers from China. The Federal Communications Commission (FCC) is advancing a related ban involving new optical modules from China, which are used to transmit data via fiber optics within data centers and connect servers, routers, and switches. Microsoft has met with officials from the White House Office of Science and Technology Policy and the National Security Council to explain the supply chain challenges that the ban could bring.John Miller, Vice President of Policy and General Counsel at the Information Technology Industry Council, stated that many cloud service providers have procured AI chips, and without optical modules, these chips will not be able to interconnect. According to Cleanview data, Microsoft currently has 35 data centers under construction in the United States, with another 134 planned. Counterpoint Research data shows that Chinese companies account for about two-thirds of the global optical module market, with Eoptolink and Zhongji Xuchuang as major suppliers.

MiniMax officially released H3-Image and open-sourced the model weights. This model can directly generate images of 4096×4096, approximately 16.8 million pixels, without additional upscaling. According to official data, the generation time is less than 10 seconds, with speeds reaching up to 5 times that of mainstream products. Text-to-image and image editing have been integrated into the same model, supporting up to 16 reference images.H3-Image follows the architecture of the H3 video model, removing the audio component while retaining the original text encoder and redesigning the decoder for images. The core image generation network has 33.1 billion parameters, requiring only 6 sampling steps to generate an image. On August 7, MiniMax researcher Kiro revealed in a Reddit AMA that the team is developing a dedicated image model based on the H3 homogenous architecture and plans to open-source the weights. They also proposed a workflow where the image model generates the first frame, which is then handed over to H3 for video generation.On November 3, Pruna AI launched P-H3-Image based on the open-source H3 video model, supporting raw images and retouching, compressing the generation network from 33.1 billion to approximately 20 billion parameters, requiring only 8 sampling steps. The minimal quantization version is about 20 GB, with peak memory usage of approximately 21 GB when generating 8 million pixel images. The official MiniMax version retains the 33.1 billion parameter generation network, with a sampling step count of 6, and supports higher resolutions and more reference images, with the original weights plus the reused text encoder totaling approximately 135 GB. There are regional restrictions on the open-source license; the United States, United Kingdom, European Union, and South Korea are not within the community license authorization regions. Companies with annual revenues exceeding 1 million dollars or those providing model hosting services to third parties must obtain separate authorization in advance.

On November 10, AI company Anthropic disclosed in its IPO roadshow video that the company believes it can achieve an adjusted operating profit margin of 35% or more in the long term, up from 15% in the third quarter of 2026, but did not provide a specific timeline for achieving this. Chief Financial Officer Krishna Rao stated that the company plans to improve its long-term cost structure by enhancing computing efficiency, increasing the proportion of directly leased chips and data centers, and expanding the share of its inference business. The ratio of computing and infrastructure expenses to revenue aims to decrease from 58% in the third quarter to 45% to 50%; the proportion of sales, marketing, and partnership expenses is expected to drop from 14% to 7% to 10%, and adjusted personnel and other operating expenses are expected to decrease from 13% to 7% to 10%.As of the end of September, there were over 11,900 enterprise customers that contributed more than $100,000 in revenue to Anthropic over the past 12 months, including over 1,900 customers contributing more than $1 million and about 250 customers contributing over $10 million. The net revenue retention rate for enterprise customers is approximately 900%, with more than half of the enterprise customers using both the Claude Platform and at least one Anthropic proprietary application. The roadshow also showcased application cases from clients such as Eli Lilly, Rakuten, Salesforce, and Walmart. The company measures computing efficiency by the ratio of revenue to computing and infrastructure expenses, which increased from 0.40 times in the first quarter of 2025 to 1.72 times in the third quarter of 2026, approximately 4.3 times higher than before, but lower than the second quarter's 2.08 times. The company stated that it used about $5 billion in short-term computing in the third quarter, with costs significantly higher than those for computing procured under multi-year agreements.In terms of actual performance, Anthropic reported third-quarter revenue of $16.956 billion, adjusted operating profit of $2.463 billion, and an adjusted operating profit margin of 15%, down from 19% in the second quarter. Under GAAP standards, the company still recorded an operating loss of $567 million in the third quarter. The adjusted metrics exclude expenses such as equity incentives and non-cash charitable donations and cannot be equated to GAAP earnings. As of the end of October, the company's annualized revenue run rate was approximately $81 billion, not the actual revenue achieved for the full year. Additionally, Anthropic disclosed long-term commitments for computing and infrastructure of about $90 billion, of which approximately $45.16 billion is expected to occur by the end of 2030, with related commitments maturing in the next 12 months amounting to about $7.07 billion. As of the end of the third quarter, the company held approximately $96 billion in cash and cash equivalents and about $36 billion in short-term investments.

On November 11, according to reports from Business Insider and Benzinga, Yorkville Ives Senior Managing Director Dan Ives stated in the latest investor report that if Jensen Huang were to be removed from his position as CEO of Nvidia, the value of Nvidia as a company and stock would significantly shrink by 25% to 30%, and it would change the company's future direction, as well as the AI GPU competitive landscape in the United States and globally, especially in China.Ives referred to the extramarital affair involving Jensen Huang and a suspected Chinese spy as Nvidia's "moment of shame." He cited Wall Street estimates that the probability of Huang being replaced by the board is about 15%, and the board's investigation may take 4 to 6 weeks, during which there will be significant pressure on the stock price. Ives stated that Huang is the core DNA of Nvidia, and he is irreplaceable; currently, no one in the management can seamlessly take over as CEO, highlighting the key person risk for Nvidia. He maintains Nvidia's "outperform" rating with a target price of $300, but on the condition that Huang continues to serve.

Artificial intelligence company Anthropic disclosed in its newly submitted S-1 filing that the seven co-founders plan to sell 3,378,378 shares each in the IPO. The preliminary price range for this IPO is set at $70 to $78 per share, which means that at the midpoint price of $74, each person will receive just under $250 million; based on the lower and upper limits of the range, each person will receive slightly more than $236 million and slightly less than $264 million, respectively. The company's IPO valuation could reach $14 billion.The seven co-founders include CEO Dario Amodei, President and Chairwoman Daniela Amodei, Chief Computing Officer Tom Brown, Head of Public Interest Jack Clark, Chief Science Officer Jared Kaplan, Chief Architect Sam McCandlish, and Head of Explainability Research Christopher Olah. After the sale is completed, each will hold approximately 3.2% of the Class A shares outstanding, totaling about 22%; each will also receive one share of Class F stock with super voting rights, giving the seven a combined control of 51.1% of the voting power.

According to Bloomberg, some investors in Anthropic have introduced expected shares to new buyers before the IPO pricing. The brokerage firm JonesTrading is representing investors expected to receive a large allocation of shares worth up to $500 million, planning to charge an additional fee of 3% to 5% based on the final IPO issue price.Insiders say that these shares are not subject to lock-up agreements, but buyers must simultaneously obtain approval from both the seller and a bank participating in the IPO. Bloomberg points out that such arrangements allow investment institutions to reduce the risk of certain allocation positions while expanding their own investor base. Jon Jefferies, head of private markets at JonesTrading, did not comment on specific transactions but mentioned that he first saw such arrangements during the SpaceX IPO, and there are currently several similar transactions related to Anthropic being pitched in the market.Additionally, another seller is promoting over $1 billion worth of existing shares in Anthropic through a special purpose vehicle (SPV), with a valuation corresponding to $14 trillion to $16 trillion for the company, and also charging fees. Insiders say that the SPV has a single-layer structure and directly holds Anthropic shares, with the transaction involving issued shares and not relying on the seller's allocation in the IPO. Anthropic's IPO seeks a fully diluted valuation of about $14 trillion, and the company along with some investors aims to raise up to $90.4 billion through the IPO and private placements combined; if this scale is reached, it will exceed SpaceX's base issuance scale of $75 billion. An Anthropic spokesperson did not immediately respond to a request for comment.

According to Gelonghui, sources have revealed that Apple is developing a Mac application named Harbor, which can turn computers equipped with Apple chips into private AI inference servers, supporting direct downloading, management, and running of open weight models without the need to upload data to the cloud.The application can also connect two or more compatible Macs via Thunderbolt 5, automatically forming a cluster and integrating unified memory to run large models that exceed the capacity of a single machine or handle more concurrent requests without manual configuration or using the command line. The application is currently in a non-public preview by Claris, a software company under Apple, and the official launch date has not yet been announced.