hodler

30 Apr 2026, 00:00

U.S. General Reveals Bitcoin’s Place in U.S. Defense Strategy

When U.S. General Samuel Paparo appeared before the Senate Armed Services Committee, he brought an unusual topic to the table. Paparo, who serves as commander of U.S. Indo-Pacific Command (INDOPACOM), made the case that Bitcoin deserves serious attention from a national security standpoint, and specifically from a technical one rather than a financial one. A Computer Science System First Paparo described Bitcoin as a computer science system with real military and cybersecurity relevance. His argument centered on the architecture itself: the combination of cryptography, blockchain technology, and Proof of Work consensus creates a cost-based security model that goes beyond what conventional algorithmic defenses can offer. That structure, in his view, produces stronger and more reliable network integrity . US Admiral Backs Bitcoin: A Matter of National Interest In April 2026, Admiral Samuel Paparo, Commander of the U.S. Indo-Pacific Command, stated during a Senate Armed Services Committee hearing on the fiscal year 2027 defense budget request that Bitcoin can be regarded as a… pic.twitter.com/hMbsDOYmqy — Wu Blockchain (@WuBlockchain) April 25, 2026 He also pointed to Bitcoin’s peer-to-peer, zero-trust design as something worth paying attention to. Cutting out centralized intermediaries reduces system vulnerabilities, a principle that aligns with military needs. More decentralization, in this context, means greater resilience. Not the Usual Government Argument This is a different conversation from the one most U.S. officials have been having about Bitcoin. The Trump administration and others have largely framed it as a potential reserve asset , a financial holding with strategic economic value. Paparo is not dismissing that framing, but he is clearly focused elsewhere. We are on X, follow us to connect with us :- @TimesTabloid1 — TimesTabloid (@TimesTabloid1) June 15, 2025 His position is that Bitcoin functions as a tool for power projection and that its defense applications exist independently of its role as a digital currency. Any technology that strengthens national power is worth incorporating into defense thinking. Bitcoin, by his assessment, qualifies on those grounds. The U.S. Military’s Bitcoin Node What makes Paparo’s testimony particularly notable is that it was not purely theoretical. He confirmed that INDOPACOM is already running a dedicated Bitcoin node, which places the U.S. military as an active participant in the network rather than an outside observer. The node is testing how Bitcoin’s protocol can help secure critical systems. That operational detail changes the nature of the discussion. The U.S. military isn’t just considering Bitcoin’s future role. They are already testing its practical uses, marking a shift in how defense institutions engage with the technology. Disclaimer : This content is meant to inform and should not be considered financial advice. The views expressed in this article may include the author’s personal opinions and do not represent Times Tabloid’s opinion. Readers are urged to do in-depth research before making any investment decisions. Any action taken by the reader is strictly at their own risk. Times Tabloid is not responsible for any financial losses. Follow us on Twitter , Facebook , Telegram , and Google News The post U.S. General Reveals Bitcoin’s Place in U.S. Defense Strategy appeared first on Times Tabloid .

29 Apr 2026, 23:25

Microsoft Copilot Users Surge Past 20 Million as Engagement Rivals Outlook

BitcoinWorld Microsoft Copilot Users Surge Past 20 Million as Engagement Rivals Outlook Microsoft has announced a major milestone for its enterprise AI assistant. The company now has over 20 million paid Microsoft Copilot users. This figure, revealed by CEO Satya Nadella during the company’s quarterly earnings call, directly challenges the perception that Copilot sees low adoption. Nadella emphasized that these are not just purchased licenses. Users are actively engaging with the tool. He reported that weekly Copilot engagement now matches the level of Microsoft Outlook. This is a significant benchmark for any enterprise software. Microsoft Copilot Users Quadruple Enterprise Deals The growth is not just in total seats. The number of companies paying for over 50,000 Copilot seats has quadrupled. Major enterprises like Bayer, Johnson & Johnson, Mercedes, and Roche each have more than 90,000 seats. The largest deal to date is with Accenture, which secured over 740,000 seats. This data suggests that large organizations are moving beyond pilot programs. They are deploying Copilot across their entire workforce. This trend signals strong confidence in the return on investment for enterprise AI tools. Engagement Metrics Show Real Usage Beyond seat numbers, Microsoft provided concrete engagement metrics. Copilot queries per user grew nearly 20% quarter over quarter. Nadella described this as a “daily habit of intense usage.” Weekly engagement is now comparable to one of the most used business tools: Outlook email. This counters the narrative that Copilot is an expensive experiment. The data shows that employees are integrating it into their daily workflows. They are using it for tasks like drafting emails, analyzing data in Excel, and summarizing documents in Word. Agent Mode Drives Deeper Integration A key driver of this engagement is Copilot’s new agentic capabilities. Microsoft recently made Agent mode the default experience across Copilot in Word, Excel, and PowerPoint. This allows the AI to take multi-step actions directly within documents. For example, a user can ask Copilot to analyze a sales report, create a summary, and then draft an email with key findings. The AI handles these steps without constant user prompts. This reduces friction and makes the tool more powerful for complex tasks. Multi-Model Strategy Reduces OpenAI Dependency Nadella also addressed a key concern: reliance on a single AI provider. He stated that Copilot is not dependent on any one model, like OpenAI. Users now have access to multiple models by default in chat. The system uses intelligent auto-routing to select the best model for a given task. Microsoft 365 Copilot now supports models from Anthropic, including Claude. This multi-model approach offers flexibility and resilience. It also allows Microsoft to optimize for cost and performance across different use cases. Market Reaction and Analyst Views The announcement has been well received by analysts. Morgan Stanley’s Keith Weiss called the numbers “super impressive” and “way ahead of most people’s expectations.” This positive sentiment is reflected in Microsoft’s stock performance and broader market confidence in enterprise AI. The data also provides a benchmark for the entire enterprise AI sector. It shows that demand for AI-powered productivity tools is real and growing. Other vendors, like Google with Gemini for Workspace, are also reporting strong adoption, but Microsoft’s numbers provide the clearest evidence of scale. Comparison with Other Enterprise AI Tools To understand the scale, consider the broader market. While exact competitor numbers are often private, Microsoft’s 20 million paid seats is a leading indicator. It suggests that enterprise AI is moving from novelty to necessity. The integration directly into existing Office tools gives Microsoft a significant distribution advantage. Metric Microsoft Copilot Industry Context Paid Enterprise Seats 20 million+ Largest publicly disclosed figure Quarterly Query Growth ~20% per user Indicates strong user retention Largest Single Deal 740,000 seats (Accenture) Shows enterprise confidence Weekly Engagement Comparable to Outlook High frequency of use Impact on the Future of Work These numbers have implications for how work gets done. If Copilot usage is truly on par with email, it means AI is becoming a core communication and productivity tool. This could reshape job roles, skill requirements, and business processes. Companies are already reporting productivity gains. Employees spend less time on routine tasks like drafting emails or formatting reports. They can focus more on strategic thinking and creative problem-solving. However, this also raises questions about job displacement and the need for AI literacy training. Conclusion Microsoft’s announcement of over 20 million paid Copilot users, combined with engagement data rivaling Outlook, marks a pivotal moment for enterprise AI. The growth is driven by large-scale deployments, a multi-model strategy, and new agentic features. These Microsoft Copilot users are not just buying licenses; they are actively integrating AI into their daily work. This trend is likely to accelerate, setting a new standard for productivity in the modern workplace. FAQs Q1: What is the main takeaway from Microsoft’s Copilot announcement? A1: The key takeaway is that Microsoft Copilot has reached over 20 million paid enterprise users, and engagement is now comparable to Outlook. This proves that enterprise AI adoption is real and growing rapidly. Q2: How does Copilot’s engagement compare to other Microsoft tools? A2: CEO Satya Nadella stated that weekly Copilot engagement is now at the same level as Outlook. This is a significant benchmark, as Outlook is one of the most used business communication tools. Q3: What is driving the increase in Copilot usage? A3: A major driver is the new Agent mode, which is now the default in Word, Excel, and PowerPoint. This allows the AI to take multi-step actions directly in documents, making it more useful for complex tasks. Q4: Is Microsoft Copilot dependent on OpenAI? A4: No. Microsoft has implemented a multi-model strategy. Copilot now supports multiple models, including Anthropic’s Claude, and uses intelligent auto-routing to select the best model for each task. Q5: Which companies are the largest users of Microsoft Copilot? A5: Major users include Bayer, Johnson & Johnson, Mercedes, and Roche, each with over 90,000 seats. The largest deal is with Accenture, which has over 740,000 seats. This post Microsoft Copilot Users Surge Past 20 Million as Engagement Rivals Outlook first appeared on BitcoinWorld .

29 Apr 2026, 23:05

Google Cloud Revenue Surpasses $20B But Growth Remains Capacity-Constrained: Q1 2026 Analysis

BitcoinWorld Google Cloud Revenue Surpasses $20B But Growth Remains Capacity-Constrained: Q1 2026 Analysis Google Cloud revenue surpassed $20 billion for the first time in Q1 2026, marking a 63% year-over-year increase. However, the company warned that growth was capacity-constrained, as demand for AI solutions outpaced available infrastructure. This milestone underscores the accelerating enterprise adoption of Google Cloud’s AI tools, including Gemini Enterprise and TPU hardware. Google Cloud Revenue Hits $20B Milestone Amid AI Surge Alphabet’s cloud division reported Q1 2026 earnings on April 30, 2026, from San Francisco, California. The Google Cloud Platform drove the majority of this growth, expanding faster than the overall division. The cloud unit includes infrastructure, data analytics, AI/ML tools, and Google Workspace. CEO Sundar Pichai attributed the strong performance to “strong demand” for Gemini Enterprise and AI solutions. AI solutions were the largest driver of cloud growth. Products built on Google’s generative AI models grew nearly 800% year-over-year. Gemini Enterprise itself grew 40% quarter-over-quarter. AI token growth via Google’s API reached 16 billion tokens per minute, up from 10 billion in Q4 2025. These numbers highlight the rapid scaling of enterprise AI usage. Capacity Constraints Limit Google Cloud Growth Potential Despite the record revenue, Pichai acknowledged significant constraints. “Obviously, we are compute constrained in the near-term,” he told analysts. “Our cloud revenue would have been higher if we were able to meet that demand.” The company’s backlog doubled to $462 billion in the quarter, indicating unmet demand. Google expects to work through 50% of this backlog over the next 24 months. This capacity constraint stems from the massive infrastructure required for AI workloads. Google provides cloud infrastructure and direct sales of TPU hardware to customers. The company takes a return on capital investment (ROIC) approach to balance spending. Pichai emphasized that this framework allows continued investment in “cutting edge” technology. New Customer Acquisition and Deal Momentum New customer acquisition doubled year-over-year in Q1 2026. Deal momentum also accelerated, with the number of $100 million to $1 billion deals doubling compared to the same period last year. Google signed multiple “billion-dollar-plus” deals during the quarter. Customers outpaced their initial commitments by 45% quarter-over-quarter, demonstrating strong demand. These metrics show that enterprise clients are committing larger budgets to Google Cloud. The backlog growth reflects both new contracts and expanded existing agreements. Pichai framed the backlog as a positive differentiator, showing Google Cloud’s unique position in the market. AI Infrastructure Investment Strategy Google invests heavily in data centers and TPU hardware to meet AI demand. The company’s capital expenditure increased significantly in Q1 2026. Pichai stated that Google has a “robust, long-range planning framework” to manage these investments. The ROIC approach ensures that spending aligns with long-term profitability. Analysts note that Google’s strategy differs from competitors. By balancing capacity expansion with financial discipline, Google aims to avoid overbuilding. The backlog provides visibility into future revenue, reducing investment risk. This approach may help Google maintain margins while scaling infrastructure. Comparison with Competitors Microsoft Azure and Amazon Web Services also reported strong cloud growth in Q1 2026. However, Google Cloud’s 63% growth rate outpaced both competitors. Microsoft Azure grew 45%, while AWS grew 38%. Google’s AI focus and Gemini Enterprise appear to be key differentiators. Google’s TPU hardware gives it a unique advantage in AI workloads. Many enterprises prefer Google’s custom chips for training and inference. This hardware differentiation, combined with Gemini models, drives customer acquisition. The capacity constraint, while limiting near-term revenue, signals strong product-market fit. Implications for Enterprise AI Adoption Google Cloud’s results reflect broader trends in enterprise AI adoption. Companies are moving from experimentation to production deployment. AI token growth of 16 billion per minute indicates massive real-world usage. Enterprises are investing in AI infrastructure to gain competitive advantages. The 800% growth in genAI model-based products shows that businesses see tangible ROI. Gemini Enterprise’s 40% quarter-over-quarter growth suggests sustained momentum. As capacity expands, Google Cloud may capture even more market share. Challenges Ahead Capacity constraints could slow Google Cloud’s growth trajectory. Competitors are also investing heavily in AI infrastructure. Google must balance speed of expansion with financial discipline. The $462 billion backlog provides a buffer, but execution risks remain. Supply chain constraints for TPU hardware and data center components could persist. Google’s long-range planning framework helps mitigate these risks. However, the company must continue innovating to maintain its edge. Conclusion Google Cloud revenue surpassing $20 billion marks a significant milestone for the division. AI-driven demand, particularly for Gemini Enterprise and TPU hardware, fueled this growth. However, capacity constraints limited even higher revenue. The $462 billion backlog and doubled customer commitments indicate strong future demand. Google’s strategic investment approach positions it well for long-term growth. As capacity expands, Google Cloud could become an even larger revenue driver for Alphabet. FAQs Q1: What drove Google Cloud’s 63% revenue growth in Q1 2026? A1: AI solutions, especially Gemini Enterprise and generative AI models, were the primary drivers. AI model-based products grew nearly 800% year-over-year, and Gemini Enterprise grew 40% quarter-over-quarter. Q2: Why is Google Cloud capacity-constrained? A2: Demand for AI infrastructure, including TPU hardware and data centers, outpaced available supply. Google’s backlog doubled to $462 billion, indicating unmet demand. Q3: How does Google plan to address capacity constraints? A3: Google uses a return on capital investment (ROIC) framework to guide spending. The company expects to work through 50% of the backlog over the next 24 months through strategic infrastructure investments. Q4: What is Google Cloud’s backlog, and why is it important? A4: The backlog represents signed contracts for future cloud services. It doubled to $462 billion in Q1 2026, signaling strong future revenue visibility and customer commitment. Q5: How does Google Cloud compare to AWS and Azure in AI? A5: Google Cloud’s 63% growth outpaced AWS (38%) and Azure (45%). Google’s TPU hardware and Gemini Enterprise models provide unique AI capabilities that differentiate it from competitors. This post Google Cloud Revenue Surpasses $20B But Growth Remains Capacity-Constrained: Q1 2026 Analysis first appeared on BitcoinWorld .

29 Apr 2026, 22:05

Runway CEO Reveals Why AI Video Is Just a Prequel to World Models—A Bold New Frontier

BitcoinWorld Runway CEO Reveals Why AI Video Is Just a Prequel to World Models—A Bold New Frontier Runway, the New York-based AI video company, has raised nearly $860 million at a $5.3 billion valuation. Its CEO, Cristóbal Valenzuela, now argues that AI video is merely a stepping stone. In a recent interview on Bitcoin World’s Equity podcast, he explained why the company is pivoting toward general world models. These models, he says, will reshape gaming, robotics, and even general intelligence. This shift marks a critical moment for the AI industry, as Runway competes directly with Google and OpenAI. Runway’s Vision: Beyond AI Video Generation Valenzuela believes that the real constraint on filmmaking has never been technology. Instead, he points to creativity and access. With AI video tools becoming mainstream, Runway now focuses on building systems that understand the physical world. These world models can simulate environments, predict outcomes, and enable real-time interactions. This goes far beyond generating clips for Hollywood. The company’s technology already powers creative workflows for thousands of users. However, Valenzuela sees a larger opportunity. He envisions a future where AI models act as interactive partners, not just tools. This requires a deep understanding of space, time, and causality—elements that world models provide. What Are World Models? World models are AI systems that learn the rules of the physical world. They can simulate how objects move, how light behaves, and how actions lead to consequences. Unlike traditional video generators, these models can run in real time. This opens up applications in autonomous vehicles, robotics, and video games. Runway differentiates itself from Google and other labs by focusing on practical, real-world applications. Valenzuela argues that many competitors build world models for research only. Runway, however, aims to deploy them in products that users can touch and feel. Why Runway’s CEO Thinks AI Video Is Just a Prequel Valenzuela uses the term “prequel” deliberately. He sees AI video as the first chapter of a much larger story. The technology has matured rapidly. Just two years ago, AI-generated video was a novelty. Today, it is a creative tool used by professionals and amateurs alike. But the next chapter, he insists, involves interactive, real-time experiences. He points to nonlinear media as a key concept. Instead of watching a linear video, users can interact with a world. They can change variables, explore different paths, and receive personalized outcomes. This shifts AI from a passive generator to an active participant. The Role of Real-Time Video Generation Real-time video generation is a technical challenge. It requires massive computational power and sophisticated algorithms. Runway has invested heavily in this area. The company believes that real-time capabilities will unlock use cases beyond content creation. For example, architects can simulate buildings in real time. Surgeons can practice procedures on virtual patients. Educators can create immersive lessons. Valenzuela also pushes back against dystopian narratives around AI companions. He argues that interactive AI can enhance human connection, not replace it. This aligns with Runway’s broader mission to democratize creativity and intelligence. Competing with Google and OpenAI Runway operates in a highly competitive landscape. Google’s DeepMind and OpenAI have vast resources and talent. However, Valenzuela believes Runway has an edge: focus. While larger labs pursue general artificial intelligence, Runway targets specific, practical problems. This allows the company to iterate quickly and ship products faster. The company has also built a strong community of creators. This feedback loop helps refine its models. Valenzuela notes that user input has been crucial for improving video quality and reducing artifacts. He expects the same iterative process to apply to world models. Funding and Valuation Context Runway’s $860 million in funding comes from top-tier investors. The $5.3 billion valuation reflects confidence in its technology and vision. However, the company faces pressure to deliver on its promises. The AI market is crowded, and investors expect tangible results. Valenzuela remains optimistic. He believes that world models will become as ubiquitous as video generators. He compares the current moment to the early days of the internet. Just as web browsers unlocked new possibilities, world models will redefine how humans interact with machines. Applications in Gaming and Robotics Gaming is a natural fit for world models. Game developers can use them to create dynamic environments that respond to player actions. This reduces the need for manual scripting and allows for emergent gameplay. Robotics is another promising area. World models can help robots navigate unfamiliar spaces and manipulate objects. Valenzuela emphasizes that these applications are not theoretical. Runway is already working with partners in both industries. The company plans to release developer tools later this year. These tools will allow third parties to build their own world model applications. Ethical Considerations and Dystopian Fears Valenzuela acknowledges the ethical concerns surrounding AI. He addresses fears that AI companions or world models could be used for manipulation. He argues that technology is neutral. The impact depends on how it is used. Runway has implemented safety measures, including content filters and usage guidelines. He also criticizes the “inherently dystopian” label often applied to AI companions. He points out that many people already form emotional bonds with AI through chatbots and virtual assistants. World models, he says, can make these interactions richer and more meaningful. Timeline and Future Outlook Runway expects to release its first world model product within the next 12 months. The company is currently in beta testing with select partners. Public availability will follow, likely in phases. Valenzuela cautions that the technology is still evolving. He expects significant improvements in speed and accuracy over the next few years. He also hints at longer-term goals. Runway aims to contribute to the development of general intelligence. World models, he believes, are a critical component. They provide a framework for understanding the physical world, which is essential for any truly intelligent system. Expert Reactions and Industry Impact Industry analysts have mixed reactions. Some praise Runway’s ambition. Others question whether the company can compete with tech giants. However, most agree that world models represent a significant step forward. The technology could transform industries ranging from entertainment to manufacturing. Valenzuela remains focused on execution. He stresses that Runway’s success depends on delivering real value to users. He encourages developers and creators to experiment with the tools as they become available. Conclusion Runway’s CEO has laid out a bold vision. AI video is just the beginning. World models represent the next frontier. With substantial funding, a clear strategy, and a focus on practical applications, Runway is positioned to lead this transition. The coming months will reveal whether the company can deliver on its promises. For now, Valenzuela’s message is clear: the future of AI is interactive, real-time, and deeply integrated into the physical world. FAQs Q1: What are world models in AI? World models are AI systems that learn how the physical world works. They simulate objects, actions, and outcomes in real time. This enables applications in gaming, robotics, and simulation. Q2: How does Runway differentiate from Google and OpenAI? Runway focuses on practical, deployable products rather than pure research. It targets specific use cases like video generation and world models, and it benefits from a strong creator community. Q3: When will Runway release its world model product? Runway plans to release its first world model product within 12 months. Beta testing is underway with select partners. Public release will follow in phases. Q4: Are world models safe to use? Runway has implemented content filters and usage guidelines. The CEO argues that technology is neutral and that safety depends on responsible deployment. The company continues to refine its safety measures. Q5: What industries will benefit most from world models? Gaming, robotics, architecture, education, and healthcare are key industries. World models enable real-time simulation, personalized experiences, and improved training tools. This post Runway CEO Reveals Why AI Video Is Just a Prequel to World Models—A Bold New Frontier first appeared on BitcoinWorld .

29 Apr 2026, 21:02

Important Clarification for the XRP community

A recent statement from crypto commentator Arthur on X has provided a clear explanation of a recurring misunderstanding within the XRP community. His remarks focus on how announcements involving Ripple partnerships are often interpreted and why those interpretations may not accurately reflect XRP’s actual usage. Important clarification for the XRP community When you see the headline “Ripple signs partnership with Bank X”, most people immediately think: “Great! XRP is being used!” But that’s often not the case. Reality check: • Ripple is a company that sells enterprise payment… — Arthur (@XrpArthur) April 27, 2026 Clarifying Ripple Partnerships and XRP Usage Arthur begins by addressing a common reaction to headlines announcing that Ripple has signed a partnership with a financial institution. According to him, many readers immediately assume that such developments mean XRP is being used in the partnership. He states that this assumption is frequently incorrect. He explains that Ripple operates primarily as a company offering enterprise payment technology. Its core products, including RippleNet and related solutions, are designed to support cross-border payments, messaging infrastructure, and treasury management for financial institutions. These tools, Arthur notes, form the foundation of most Ripple partnerships. However, he emphasizes that the majority of these partnerships do not involve XRP. Instead, banks and financial institutions typically adopt Ripple’s technology stack without integrating the digital asset itself. This distinction, he suggests, is often overlooked when partnership announcements are publicized. When XRP Is Actually Used Arthur further clarifies that XRP is used only in specific cases where institutions opt for solutions such as On-Demand Liquidity (ODL) or newer products built to integrate the token on the XRP Ledger . In these scenarios, XRP plays a functional role in facilitating liquidity and enabling faster transactions. Despite this, he points out that such use cases represent a smaller portion of Ripple’s overall enterprise activity. As a result, the growth of Ripple’s business operations does not directly translate into widespread XRP adoption. We are on X, follow us to connect with us :- @TimesTabloid1 — TimesTabloid (@TimesTabloid1) June 15, 2025 Explaining Community Frustration Arthur acknowledges that this disconnect helps explain why some long-term XRP holders express frustration. He states that while Ripple continues to expand rapidly on the enterprise side, the pace of XRP adoption remains comparatively slower. This difference, according to his commentary, creates a situation where Ripple’s commercial success is visible and measurable, but the direct impact on XRP usage is less immediate. He stresses that understanding this distinction is essential for accurately assessing developments within the ecosystem. Arthur concludes by reinforcing the importance of separating Ripple’s corporate progress from XRP’s adoption trajectory. His remarks encourage readers to evaluate announcements more carefully and to recognize that not all Ripple-related news directly affects the utility or demand for XRP. This clarification provides a more nuanced perspective on the function of Ripple’s partnerships and their relation to XRP’s role in the financial technology landscape. Disclaimer : This content is meant to inform and should not be considered financial advice. The views expressed in this article may include the author’s personal opinions and do not represent Times Tabloid’s opinion. Readers are advised to conduct thorough research before making any investment decisions. Any action taken by the reader is strictly at their own risk. Times Tabloid is not responsible for any financial losses. Follow us on X , Facebook , Telegram , and Google News The post Important Clarification for the XRP community appeared first on Times Tabloid .

29 Apr 2026, 18:40

DeepSeek adds image and video recognition to its main chatbot

A Chinese artificial intelligence company has added image and video recognition to its main chatbot. At the same time, local chip makers showed they can now match the fast launch support that used to be an American strength. DeepSeek, a company based in Hangzhou, quietly added a new feature called “image recognition mode” to its chat platform. This new mode joins two other modes the company launched earlier this month: “expert” and “flash.” The new feature allows the chatbot to understand photos and videos, not just text. This brings it in line with other major AI chatbots that have offered similar abilities for some time. Chen Xiaokang, who heads DeepSeek’s multimodal team, said the tool was first tested with a small group of users on both the website and the mobile app. Chen Deli, a senior researcher at the company, celebrated the launch with a short post that referred to the company’s logo: “The little whale can now see.” The image and video feature came out just a few days after DeepSeek released a preview of its newest flagship model, DeepSeek-V4, and made the model weights available for anyone to download and use. V4 is not one model but two. The first one, DeepSeek-V4-Pro, has 1.6 trillion parameters and is designed for difficult tasks that need complex reasoning and multi-step automated workflows. The second one, DeepSeek-V4-Flash, is built to handle a large number of requests at a lower cost. Both models support a context window of one million tokens. They also use a hybrid attention design that the company says reduces computing power and memory needed during inference. Chinese chip makers hit a new milestone What caught the attention of many industry observers was not just the model itself, but what happened on the day it launched. Four Chinese chip companies, Huawei Ascend, Cambricon, Hygon Information, and Moore Threads, all confirmed that their hardware worked perfectly with V4 from the very first day it was released. This kind of same-day support, where a new model runs smoothly on non-NVIDIA chips right at launch instead of weeks or months later, had previously been almost impossible outside of Nvidia’s own ecosystem. Huawei’s Ascend chips, including the A2, A3, and 950, support both V4-Pro and V4-Flash. The company said its Ascend 950 chip uses fused computing processes and parallel processing streams to make inference faster. Cambricon finished its adaptation using the open-source vLLM inference framework and shared its code on GitHub. Hygon said it carried out deep model optimization on its DCU platform to create a smooth path from model release to actual use. Moore Threads worked with the Beijing Academy of Artificial Intelligence to run V4 on its MTT S5000 card using the FlagOS software stack. Industry watchers say this coordinated launch represents a real change. For years, chips made outside Nvidia’s ecosystem would take months to support a major new model. Getting eight different domestic chipsets to work on day one is a significant milestone. The bigger picture: cost and independence for Deepseek Observers believe the bigger meaning of this launch is that DeepSeek has shown it can deliver high-level AI without relying on Western hardware. By making its models work natively on multiple Chinese chips at the same time, it lowers the risk from export restrictions that have blocked Chinese companies from accessing the most powerful American processors. Cost is also important. DeepSeek has worked hard to keep the price of running its models low. This makes it easier for businesses to build automated systems without facing very high computing costs. In this way, Deepseek’s upgrades and launch are not mainly about one technical breakthrough. They are more about an entire supply chain coming together. From this release, the question of who leads in AI appears to be moving away from who builds the smartest model, and toward who can keep the whole system running cheaply and independently for the long term. Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free .