News linked to both this project and an event.
Google has begun rolling out Gemini Spark to Google AI Pro users outside the United States. Spark is a personal AI agent that runs 24/7 in the background, capable of handling heavy tasks based on user instructions. According to the test account Testing Catalog, the product's next major upgrade will support true MCP (Model Context Protocol), with related features already in development.
According to the UK's Financial Times (FT), Google DeepMind has disbanded the development team behind the AI model "AlphaFold," which won the Nobel Prize for protein structure prediction. Most of the paper's co-authors have been reassigned to departments such as enzyme design, nuclear fusion, genomics, and Gemini-related projects within the past year; some members transferred to Isomorphic Labs, a new drug R&D subsidiary under Alphabet, while another quarter of full-time employees have left. DeepMind Vice President of Research Pushmeet Kohli stated that the company's strategy has shifted from focusing on single scientific challenges to building Gemini-based systems to support scientists and drive automation of scientific processes, while competing in frontier AI agent development with OpenAI, Anthropic, and others.
据 e27 报道，谷歌发布《2026 年东南亚 Gemini 报告》显示，东南亚地区每周一专业生产力提示词使用量较平日上升 30%，AI 已深度融入职场节奏。 报告指出，菲律宾因拥有庞大的 BPO（业务流程外包）产业，客服相关提示词使用量是东南亚平均水平的近 3 倍，AI 被用于缓解客服人员的沟通压力与重复性文档工作，被称为"倦怠屏障"。菲律宾也是区域内唯一女性用户请求量超过男性的市场，写作与求职类需求尤为突出。 新加坡则以全球最高的 Gemini 人均使用率领跑，60% 的提示词来自桌面端，用户每日平均发起近 10 次交互，主要用于代码调试、数据清洗、演示文稿摘要及深度研究等知识型工作。 周末期间，东南亚整体创意与休闲类提示词上升 12%，AI 使用场景从职场延伸至个人理财、旅行规划与健康饮食等日常生活领域。
Cryptocurrency exchange Gemini has donated approximately $10 million worth of Bitcoin to MAGA Inc., a super PAC supporting former US President Donald Trump. According to a report filed with the US Federal Election Commission (FEC), Gemini co-founders Cameron Winklevoss and Tyler Winklevoss completed the donations in two installments on June 19, each worth over $5 million. This donation occurred about three weeks after the CFTC and Gemini jointly filed a motion with the US District Court for the Southern District of New York, seeking to vacate the $5 million settlement agreement reached in January 2025. The settlement stemmed from the CFTC's earlier allegations that Gemini had made false or misleading statements. MAGA Inc. can use these funds to support independent political expenditures for Donald Trump. Previously, the Winklevoss brothers each donated $1 million to Donald Trump's 2024 campaign and contributed $21 million worth of Bitcoin to the Digital Freedom Fund PAC to support the crypto policy direction of the Trump administration. Currently, the court has not yet ruled on the motion to vacate the settlement between the CFTC and Gemini. Meanwhile, CFTC Chairman Michael Selig remains the sole commissioner of the agency. As of June 30, MAGA Inc. has reported receiving over $397 million in funds.
B.AI API has officially launched two models: Gemini 3.6 Flash and Gemini 3.5 Flash-Lite. Among them, Gemini 3.6 Flash, as an upgraded version of 3.5 Flash, significantly improves output quality while maintaining the same pricing, with Token consumption reduced by approximately 17% on average; especially in complex DeepSWE scenarios, the reduction can reach up to 65%, greatly enhancing development efficiency and cost-effectiveness. Gemini 3.5 Flash-Lite delivers ultra-fast output of up to 350 tokens/s and ultimate cost-effectiveness, precisely meeting the demands of high-concurrency, high-real-time tasks such as document batch processing and Agentic intelligent retrieval. Currently, official API interfaces for both models are available. Developers can log in to the B.AI platform immediately to seamlessly access and experience them, enjoying the performance benefits brought by the new generation of models.
According to CryptoSlate, U.S. Federal Election Commission documents show that Cameron Winklevoss and Tyler Winklevoss donated a total of Bitcoin worth $10.018 million to the pro-Trump super political action committee MAGA Inc. on June 19, 2026. Relevant records also show that this batch of Bitcoin was subsequently sold through Gemini; the appearance of the Gemini name in the documents does not mean Gemini made a separate donation.
Google releases next-generation Gemini Flash series models, including Gemini 3.6 Flash, Gemini 3.5 Flash-Lite, and Gemini 3.5 Flash Cyber for cybersecurity scenarios.
sources familiar with the matter have revealed Google is developing a new server chip, internally codenamed "Frozen v2," which will directly integrate the architectural design of the Gemini AI model to enhance AI model operational efficiency and alleviate current computing power supply pressures. Google aims to optimize the inference process of the Gemini model through this new chip, enabling it to provide AI services to users at lower costs and higher efficiency. Industry insiders believe that if Frozen v2 successfully launches, it will further strengthen Google's independent capabilities in the AI infrastructure sector and become an important bargaining chip in its competition with Nvidia and AMD for the AI computing ecosystem. (The Information)
Apple is shifting its reliance to NVIDIA GPUs to support its artificial intelligence operations, as the company's self-developed M2 Ultra chip has failed to meet expectations when running advanced AI workloads.Sources indicate that Apple previously attempted to use its own servers to run Google's Gemini model to power AI capabilities for the new Siri. However, the performance of its chips fell short of requirements, ultimately forcing the company to turn to NVIDIA GPUs deployed on Google Cloud. Additionally, the development of Apple's next-generation AI server chip, "Baltra," has been delayed, while the M7 Ultra chip, which is expected to rival the performance of NVIDIA's Blackwell architecture, is unlikely to be launched before 2029 at the earliest. (Coin Bureau)
crypto trading platform Gemini has announced the launch of zero-commission stock trading services in most U.S. states, further expanding its traditional financial business and moving toward the goal of becoming an "all-in-one financial super app."Gemini stated that the service is currently not available to users in Alabama, Arkansas, Illinois, Massachusetts, Texas, Puerto Rico, Washington D.C., and Guam. Real-time market data services will be provided by Nasdaq. (The Block)
Tech blogger Leo has revealed that OpenAI may make GPT-5.6 available to the public between July 7 and July 9, with the earliest possible date being July 7. The new model's plan usage quotas are said to be more generous, and OpenAI is further strengthening its safety strategies ahead of the launch.Additionally, according to sources, Google DeepMind has tentatively scheduled the release of Gemini 3.5 Pro for July 17. Another tech blogger, Astro Polo, stated that Gemini 3.5 Pro will support a 2 million token context window, doubling the 1 million token context window currently supported by Claude Sonnet 5, Claude Opus 4.8, and Claude Fable 5. This makes it more suitable for handling large codebases, lengthy documents, and long conversations.Note: The above information is based on market rumors and has not yet been officially confirmed by OpenAI or Google DeepMind.
According to Meituan's official release, Meituan has officially launched the new generation large model LongCat-2.0 and open-sourced it simultaneously. The model features a total of 1.6T parameters, making it the industry's first trillion-parameter model to complete full-process training and inference on a domestic computing cluster of 50,000 cards. It natively supports 1M ultra-long context and focuses primarily on code understanding, generation, and execution in Agentic Coding scenarios. Technically, LongCat-2.0 adopts the LongCat Sparse Attention (LSA) sparse attention mechanism, reducing long text computation complexity from quadratic to linear; achieves token-level dynamic activation (33B~56B) via a zero-computation expert mechanism; and introduces the MOPD architecture to fuse three sets of expert capabilities: Agent, Reasoning, and Interaction. In terms of training efficiency, the team spent three years overcoming challenges in adapting to domestic computing power, reducing the monthly average daily failure rate by over 70%, increasing training MFU by 1.5 times, and achieving steady-state daily throughput exceeding 1T tokens/day. In terms of performance evaluation, LongCat-2.0 achieved a score of 59.5 on SWE-bench Pro, surpassing Gemini 3.1 Pro (54.2), GPT-5.5 (58.6), and Claude Opus 4.6 (57.3); and achieved a score of 79.9 on BrowseComp.
Tech blogger @synthwavedd has stated that several highly anticipated cutting-edge AI models have adjusted their release schedules. Among them, GPT-5.6, originally slated for a near-term release, has been postponed to mid-July, while Google DeepMind has canceled the planned late-June launch of Gemini 3.5 Pro due to dissatisfaction with the model’s current performance.Meanwhile, preparations for the launch of OpenAI’s new-generation bidirectional voice model, Bidi, are progressing on the ChatGPT platform, with availability to users potentially coming as early as this week. Reports indicate that Bidi supports full-duplex voice interaction, allowing users to speak simultaneously with the model and interrupt the conversation at any time, marking a significant upgrade over existing voice modes.Additionally, Anthropic has granted early access to its Claude Sonnet 5 model for select enterprise clients. As the release progress of flagship models Mythos 5 and Fable 5 has slowed, Claude Sonnet 5 is seen as an interim product solution for Anthropic to alleviate competitive pressure.
the U.S. Department of Justice stated in a press release that a 47-year-old resident of Newcastle, Washington, Geoffrey K. Auyeung, has been sentenced to 5 years in prison for conspiracy to commit money laundering.Geoffrey K. Auyeung assisted overseas scammers in transferring nearly $100 million in investment fraud proceeds through bank accounts and cryptocurrency exchanges. The scammers deceived victims into investing in the oil and gas industry, luring them to transfer funds into so-called escrow accounts. To facilitate this, Geoffrey K. Auyeung established at least nine entities to receive funds, which were then transferred overseas or exchanged for cryptocurrencies such as Bitcoin, Ethereum, USDT, and USDC via exchanges like Gemini, Coinbase, and BitStamp. Most of these cryptocurrencies were subsequently sent to Binance accounts controlled by individuals in Nigeria and Russia.Geoffrey K. Auyeung opened at least 81 bank accounts across 24 financial institutions and 19 accounts across 8 cryptocurrency exchanges, receiving a total of $97.1 million in wire transfers and deposits. Through his involvement in the scheme, he earned at least $4 million in commissions and received an additional $400,000 in commissions through accounts under his wife's name between August 2024 and December 2025. Auyeung pleaded guilty in February of this year. He will forfeit approximately $2.3 million seized from bank accounts and his home, an Audi SQ8, and has agreed to forfeit approximately $7.1 million worth of cryptocurrency. (The Block)
Samsung has officially announced the advancement of its "AI Big Transformation," integrating artificial intelligence across all businesses of its affiliated companies and establishing new AI-dedicated organizations for each affiliate. This plan is a specific execution plan proposed after Samsung Electronics Chairman Lee Jae-yong's New Year's address this year, where he emphasized changing work methods and organizational DNA. Samsung will formally introduce external generative AI services such as Gemini, ChatGPT, and Claude at each affiliate within this month, and plans to apply AI to business areas including software, marketing, development, and manufacturing, covering eight core processes: development, procurement, manufacturing, logistics, marketing, sales, service, and operational support. Additionally, Samsung will conduct a two-day hands-on intensive AI training session for approximately 50 affiliate presidents within the month, and plans to provide phased training for around 2,300 affiliate executives, with AI training for all employees expected to be completed within the year. The newly established AI-dedicated organizations will be responsible for formulating AX (AI Transformation) advancement strategies for each company, managing data and model operations, and cultivating AI talent. (etnews)
据 The Block 报道，Gemini 推出 AI 驱动的个性化市场信息流 Command Center ，可基于用户偏好提供实时洞察、信号、摘要与预测，该服务已接入其预测市场，并由 Elon Musk 旗下 xAI 开发的大语言模型 Grok 提供支持。报道称，Gemini 正持续从现货加密货币交易扩展至衍生品、预测及其他金融服务。其一季度服务和利息收入同比增长超 120% 至 2450 万美元，占总营收近一半。
According to CoinDesk, cryptocurrency platform Gemini has announced the launch of its “Command Center,” an AI-powered intelligence layer built into its prediction markets platform. This feature integrates the SpaceXAI model to deliver real-time market summaries, sentiment analysis, and personalized signals across cryptocurrency, sports, commodities, economics, and politics—tailored to users’ holdings, watchlists, and prediction activities.
FTSE Russell announced the preliminary inclusion list for the Russell 3000 Index, which includes crypto-related companies such as Sharplink, Forward Industries, Gemini, Galaxy Digital, and Bitmine Immersion Technologies. Among them, Sharplink—valued at approximately $1.2 billion—and Forward Industries—valued at roughly $350 million—may qualify for inclusion in the Russell 2000; Bitmine and Galaxy Digital—which has a market capitalization of approximately $11.55 billion—may meet the eligibility threshold for the Russell 1000. The Russell 3000 tracks the 3,000 largest U.S.-listed companies by market capitalization. The final list will be updated on June 5, June 12, and June 18, and will take effect after the U.S. market close on June 26.
B.AI Platform announced that, following its initial integration on the API side, Gemini 3.5 Flash is now officially available to Web Chat users on the platform. As Google’s next-generation high-performance model, Gemini 3.5 Flash delivers fast, stable performance comparable to leading models—at a significantly lower cost—while demonstrating enhanced capabilities in code understanding, complex task execution, and agent-based application scenarios, enabling highly efficient and seamless user interactions. Users can currently experience this model directly via both B.AI’s API and Web Chat channels—try it now!
amidst the US earnings season and the continuous advancement of AI narratives by tech giants, according to official sources from Bitget, the platform has recently added 10 new US stock contract targets. As a result, Bitget's Universal Exchange (UEX) now supports 263 US stock tokens and 81 US stock contracts. The total number of US stock-related assets tradable across the entire platform has exceeded 340.This week enters a period of密集 earnings reports and macroeconomic data releases. Through continuous expansion, Bitget aims to provide users with a wide range of cross-market asset allocation channels.3 Key Focus Targets This Week:• NVDA (NVIDIA): Will release Q1 earnings after the US market closes on May 20. The implied volatility from the options market suggests a potential single-day swing of over 7.5%. As the core of AI computing power, its performance will directly impact the overall valuation trend of the tech sector.• GOOGL (Google): Recently, at its I/O conference, Google intensively released multiple new products such as Gemini 3.5 and AI agents. The market is closely monitoring the subsequent pace of commercial implementation and revenue monetization capabilities.• WMT (Walmart): Will release Q1 earnings before the market opens on May 21. As a retail giant, Walmart's resilience during the consumer downturn cycle is under significant scrutiny. The earnings results and guidance will directly influence investors' assessment of full-year growth expectations and the stock's short-term performance.