Industry Briefing
A single destination for timely, editor-curated robotics news from around the world.
DeepSeek Unveils DSec Sandbox Infrastructure for Large-Scale Agent Training
DeepSeek has introduced the DSec platform, a sandbox infrastructure tailored for extensive agent training. This innovative system integrates various sandbox types, including function-call, container, microVM, and full-VM sandboxes, while effectively managing their lifecycle alongside reinforcement learning tasks. The significance of DSec lies in its ability to support large-scale operations, with a production-scale unit comprising approximately 160 nodes. It can handle around 3 million sandboxes daily and accommodate over 380,000 concurrent sandboxes, showcasing its robust capabilities in the field of agent training. Looking ahead, the DSec platform's ability to create more than 5,000 sandboxes per second positions it as a critical tool for developers and researchers in the AI domain. No further timeline was disclosed at the time of publication.
News Feed
DeepSeek Expands Engineering Team by 150 as ByteDance Accelerates Video Generation
DeepSeek has announced a significant expansion of its engineering team, adding 150 new engineers to enhance its capabilities in AI and robotics. This move reflects the growing demand for advanced AI solutions in various sectors, particularly in the development of large models that go beyond traditional parameter counts. The expansion is crucial as it positions DeepSeek to better compete in the rapidly evolving AI landscape, where companies are increasingly focusing on innovative applications of technology. By bolstering its workforce, DeepSeek aims to accelerate the development of its products and services, particularly in the realm of video generation, where ByteDance is also making strides. Looking ahead, industry observers will be keen to see how this recruitment drive impacts DeepSeek's product offerings and market position. No further timeline was disclosed at the time of publication.
Robotics Automation AI
DeepSeek Implements New API Pricing Structure with Increases Up to 1,100%
DeepSeek has introduced a new API pricing structure that took effect at midnight on August 17. The updated pricing features peak and off-peak rates for its V4-Flash and V4-Pro models, establishing a 1:2 ratio. Notably, the peak-hour cache-hit input for the V4-Pro has seen an increase of 1,100% compared to previous rates. This significant price adjustment is crucial for users of DeepSeek's API services, as the overall increases vary between 50% and 1,100%, depending on the specific model, token type, and time slot. Such changes could impact budgeting and operational strategies for businesses relying on these services. Looking ahead, it will be important to monitor how these pricing changes affect user adoption and overall market dynamics. No further timeline was disclosed at the time of publication.
Chinese LLMs Dominate Global Token Usage for Fifteen Consecutive Weeks with DeepSeek-V4-Flash Leading
Chinese large language models (LLMs) have achieved a significant milestone by surpassing 34.25 trillion weekly tokens for the first time, according to OpenRouter data. This marks the fifteenth consecutive week that Chinese LLMs have led global token usage, with the top four positions occupied by Chinese models. The rise of DeepSeek-V4-Flash is particularly noteworthy, as its official release propelled it to the number one spot with an impressive 570 percent growth week-on-week. This surge highlights the increasing adoption and performance of Chinese LLMs in the global market. Looking ahead, the continued dominance of Chinese LLMs in token usage will be crucial to monitor, especially as they maintain their lead in the industry. No further timeline was disclosed at the time of publication.
Unitree Robotics Prices IPO at RMB150.80, DeepSeek Invests RMB141 Million
Unitree Robotics has announced the pricing of its IPO on the Shanghai STAR Market at RMB150.80 per share, with a total of 40.45 million shares to be issued. As part of this offering, Hangzhou DeepSeek AI Basic Technology Research has been allocated 933,900 shares, amounting to approximately RMB141 million, which will be subject to a 36-month lockup period. This investment from DeepSeek highlights the strategic importance of collaboration in the fields of large models, embodied intelligence, and robotics development. Unitree's decision to select strategic investors like DeepSeek reflects its commitment to fostering partnerships that can enhance its technological capabilities and market position. Looking ahead, stakeholders should monitor the development of Unitree's partnerships and the impact of this IPO on its growth trajectory. No further timeline was disclosed at the time of publication.
News Feed
Unitree Sets IPO Valuation at $9 Billion with DeepSeek's $20.9 Million Investment
Unitree has announced its initial public offering (IPO) with a valuation set at $9 billion. This significant milestone highlights the company's growth and potential in the robotics sector. DeepSeek has invested 141 million yuan (approximately $20.9 million) in Unitree, which is subject to a three-year lock-up period. The investment from DeepSeek underscores the confidence in Unitree's business model and future prospects. As the robotics industry continues to expand, Unitree's IPO could attract further investments and partnerships, enhancing its market position. The valuation reflects the increasing interest in innovative robotics solutions and the potential for substantial returns. Investors and industry watchers should keep an eye on Unitree's performance post-IPO, particularly how it leverages the funding from DeepSeek to drive growth. No further timeline was disclosed at the time of publication.
Artificial Intelligence News Robotics China IPO robotics
DeepSeek and State Oil Giant Support Unitree Robotics' $900 Million IPO
Chinese AI developer DeepSeek and a state oil giant are backing Unitree Robotics' initial public offering, which aims to raise 6.1 billion yuan ($900 million) in Shanghai. This significant investment highlights the growing confidence in the humanoid robotics sector, as the company prepares for its market debut. The backing from prominent investors like DeepSeek underscores the increasing interest and potential in the humanoid robotics market. Unitree Robotics is positioned to capitalize on this trend, aiming to enhance its production capacity to meet rising demand for humanoid robots, which are expected to play a crucial role in various applications. Looking ahead, stakeholders will be monitoring the IPO's performance closely, as it could serve as a barometer for investor confidence in the emerging robotics sector. No further timeline was disclosed at the time of publication.
DeepSeek V4-Flash Dominates Global Token Consumption as Chinese Models Excel
DeepSeek V4-Flash has emerged as the leader in OpenRouter's weekly rankings, achieving a remarkable consumption of 7.1 trillion tokens. In a significant trend, Chinese models have secured nine of the top ten positions globally, indicating a strong performance in the AI token landscape. Overall, global weekly AI token usage has surpassed 56.8 trillion, highlighting the growing demand for AI technologies. This development is crucial as it underscores the competitive landscape within AI token consumption, particularly the dominance of Chinese models. The impressive figures reflect a shift in developer preferences and the evolving price-performance dynamics that are influencing market choices. The ability of DeepSeek V4-Flash to lead the rankings demonstrates its effectiveness and appeal among developers. Looking ahead, it will be important to monitor how these trends evolve and whether DeepSeek V4-Flash can maintain its lead amid increasing competition. Additionally, the performance of Chinese models in the AI token market will be a key area to watch as developers continue to seek optimal solutions for their needs. No further timeline was disclosed at the time of publication.
Moonshot AI Unveils K3 Report Detailing Advanced Technologies for DeepSeek
Moonshot AI has released a comprehensive 47-page technical report on the K3 project, highlighting several innovative technologies. Key features include the KDA+Gated MLA hybrid attention mechanism, AttnRes cross-layer retrieval, and Stable LatentMoE, which utilizes 896 experts. Additionally, the report discusses the implementation of 51.2 million reinforcement learning sandboxes. The significance of this report lies in its potential to redefine the capabilities of DeepSeek through advanced AI methodologies. The introduction of hybrid attention and cross-layer retrieval could enhance the efficiency and effectiveness of AI systems, making them more adept at handling complex tasks. The large number of reinforcement learning sandboxes indicates a robust framework for training and testing AI models. Looking ahead, industry stakeholders should monitor how these technologies will be integrated into existing systems and their impact on AI development. The advancements presented in the K3 report may set new benchmarks in AI performance and scalability. No further timeline was disclosed at the time of publication.
Technology
Apple's AI Features Approved in China; DeepSeek Prepares for IPO
Apple's 'Apple Smart' AI model has completed its registration in China as of July 8, with Alibaba and Baidu providing support for text and image generation capabilities. This collaboration enhances the domestic version of Apple's AI suite and Siri functionality, marking a significant step in AI integration for Apple devices. The approval of Apple's AI features is crucial as it indicates the company's commitment to enhancing user experience in China, a key market for Apple. With Alibaba's Qianwen and Baidu's AI search capabilities, the partnership aims to create a competitive edge in the rapidly evolving AI landscape, which is essential for maintaining relevance in the tech industry. Looking ahead, the focus will be on how effectively these AI capabilities are integrated into Apple's product lineup and the potential impact on user engagement. No further timeline was disclosed at the time of publication.
Apple price hike sparks surge in Sam's Club purchases; DeepSeek launches major hiring; gold falls below $4000 again.
OpenAI has officially announced the launch of its new flagship AI model, GPT-5.6, on June 27, 2023. However, due to restrictions from the U.S. government, only a select group of "trusted partners" have access to this advanced model, which includes three variations: Sol, Terra, and Luna, each designed for different applications. The introduction of GPT-5.6 aims to enhance AI reasoning capabilities and streamline complex tasks. In related news, on June 26, the price of gold fell below $4,000 per ounce, influenced by several investment banks, including Goldman Sachs, lowering their price targets for the precious metal. This decline has led to a corresponding drop in gold jewelry prices across various retail outlets in China. Meanwhile, in the tech sector, Asian stock prices related to Apple's supply chain experienced significant declines following the company's announcement of price increases for its MacBook and iPad products. This has raised concerns about the potential impact of rising semiconductor costs on consumer demand and overall tech spending. In the automotive industry, Tesla has shifted its Full Self-Driving (FSD) sales model to a subscription basis globally, with plans to phase out the one-time purchase option by June 30. Additionally, Volkswagen is reportedly planning to cut up to 100,000 jobs as part of a restructuring effort aimed at reducing costs by €11 billion by 2030.
Smart Hand valued at $1 billion after 5 months, achieves profit in first quarter; DeepSeek plans to double all departments.
On June 26, 2026, Wuliangye announced the removal of Zeng Zongqin from his position as chairman due to his inability to fulfill his duties. The decision was made during the company's annual shareholder meeting, where a proposal to appoint Deng Min as the new chairman was approved. Zeng's departure, effective immediately, will not affect the company's operations, as he held no shares and had no outstanding commitments. In Yantai, JD Pharmacy has established a new company, JD Pharmacy (Yantai) Co., Ltd., with a registered capital of 100,000 RMB. The company, fully owned by JD Pharmacy (Qingdao) Chain Co., Ltd., will engage in retailing pharmaceuticals and providing medical device services. SpaceX, led by Elon Musk, plans to launch a new Starlink mobile service in the U.S. and develop its own ground mobile network, aiming to compete with major carriers like Verizon, AT&T, and T-Mobile. He Xiaopeng announced on June 25 that XPeng Motors is set to expand globally, with the second-generation VLA model expected to enter international markets by 2027, coinciding with new EU regulations on autonomous driving. Meanwhile, Lingqiao Shou, a company spun off from Zhiyuan Robotics, has achieved a valuation of $1 billion within just five months of its establishment, following significant funding rounds. Beijing Tongrentang Medical and Health Investment Co. plans to raise up to HKD 671.6 million through an IPO in Hong Kong, with shares priced between HKD 5.48 and HKD 6.21, expected to begin trading on July 7. Lastly, DeepSeek announced plans to double its workforce across all departments, while Apple reported a surge in memory demand due to the rapid expansion of AI data centers, presenting challenges for the electronics industry.
Unitree's IPO on Shanghai's Star Market Highlights Investor Interest in AI Robotics
Unitree has launched its IPO on Shanghai's Star Market, pricing shares at 150.8 yuan each. The company aims to sell 40.45 million new shares to raise 6.1 billion yuan, exceeding its initial target of 4.2 billion yuan. DeepSeek, which invested 141 million yuan, will collaborate with Unitree on AI models and embodied-intelligence technology. This IPO is significant as it positions Unitree as a key player in China's AI robotics sector, often referred to as the first humanoid stock from mainland China. The collaboration with DeepSeek, which includes a three-year lock-up period, will enhance Unitree's capabilities in AI and robotics, potentially influencing the market dynamics for similar companies. Investors should monitor how Unitree's IPO performance impacts valuations in the embodied-intelligence sector. According to Zheng Hualiang, founder of Leap VC, this IPO could set a valuation benchmark for other companies in the field. No further timeline was disclosed at the time of publication.
DeepSeek to launch V4 in mid-July with new peak-time API pricing
The DeepSeek team has announced that the official release of DeepSeek V4 is set for mid-July. This new version will enhance the existing preview release with additional features and performance improvements. Notably, it will include a 1-million-token context window across the entire model lineup, promising stronger performance capabilities. The updates aim to provide users with a more robust and efficient tool for their needs.
News Feed
Microsoft Can't Afford Unlimited Token Either — Enter DeepSeek
Microsoft has announced a shift in its pricing strategy for Copilot Cowork, moving to a usage-based model in response to rising operational costs. This decision comes as the company seeks to enhance affordability for users while managing expenses effectively. To support this transition, Microsoft is also adopting DeepSeek V4, an open-source alternative that promises to deliver cost-effective solutions for its services. The changes are set to take effect immediately, reflecting the company's commitment to adapting its offerings in a rapidly evolving market. By leveraging open-source technology, Microsoft aims to maintain competitive pricing while continuing to provide robust features to its users.
AI
DeepSeek Raises $7.4 Billion in Historic Series A: Tencent Leads, CATL Crosses Over, Alibaba and ByteDance Sit Out
Chinese AI startup DeepSeek has successfully raised around 51 billion yuan in a Series A funding round, achieving a valuation of 400 billion yuan. The funding, which took place recently, saw significant investment from tech giant Tencent, contributing 10 billion yuan, while battery manufacturer CATL has also joined as a strategic investor. This substantial financial backing highlights the growing interest and confidence in DeepSeek's innovative technology and its potential impact on the AI landscape.
AI
DeepSeek Website Hits 541M Monthly Visits in May as Baidu's AI Products Stage a Comeback
In May, DeepSeek achieved a remarkable milestone by recording 541 million monthly visits, solidifying its position as the leading search engine in China. This achievement comes amid a notable resurgence of Baidu's AI products, which encompass search functionalities, long-form content, and conversational AI. The recovery of Baidu's offerings indicates a competitive landscape in the Chinese tech market, where both companies are vying for dominance. As consumer engagement with AI-driven services continues to grow, the performance of these platforms reflects the increasing reliance on advanced technology in everyday online interactions.
AI
Who could gain from DeepSeek’s V4 with China chips poised for stronger demand?
DeepSeek, a start-up based in Hangzhou, has unveiled its latest artificial intelligence model, the V4 series, which analysts believe could significantly impact the stock market across the tech industry, from chip manufacturers to large language model developers. This launch, described as a major milestone for the company, positions DeepSeek's platform as the most powerful open-source alternative capable of competing with established US rivals. The breakthrough is expected to increase demand for computing power and promote wider commercial adoption of AI technologies. As the industry anticipates the effects of this advancement, investors are likely to reassess their positions in related stocks.
From DeepSeek to DeepRoute: Why a Top AI Researcher Bet on the Physical World
At the 2026 Beijing Auto Show, DeepRoute.ai announced a strategic transformation from being an Advanced Driver Assistance Systems (ADAS) supplier to focusing on developing Physical AI infrastructure. This shift is underscored by the integration of a unified foundation model and extensive real-world data to enhance AI applications in the physical realm. The company also highlighted its commitment to innovation with the recent addition of Ruan Chong, a former scientist from DeepSeek, to its team. This move reflects DeepRoute.ai's ambition to leverage AI technology to create more effective solutions for real-world challenges.
News
DeepSeek takes step away from Nvidia amid export curbs on China
DeepSeek, a Hangzhou-based company, has introduced its latest AI model, the V4 Pro, with significant price reductions aimed at intensifying competition in the artificial intelligence sector as China seeks to rival the United States. Announced last week, the company is offering developers a remarkable 75 percent discount on the new model. Additionally, fees for input cache hits across its AI platforms have been drastically reduced to just 10 percent of the original price, benefiting users who regularly make similar or repeat requests. This strategic pricing move is designed to attract more developers and enhance the accessibility of AI technology in a rapidly evolving market.
DeepSeek launches V3.2 models with integrated reasoning tool use
DeepSeek has announced the launch of its V3.2 and V3.2-Speciale models, now available across web, app, and API platforms. This latest version introduces built-in reasoning capabilities for agent tasks and marks the first instance of the model supporting tool calls in both reasoning and non-reasoning modes. According to the company, V3.2 achieves results comparable to GPT-5 on public reasoning benchmarks, while also optimizing output length and reducing computational costs. The release aims to enhance user experience and efficiency in various applications, reflecting DeepSeek's commitment to advancing AI technology.
News Feed
MIT Technology Review releases 2025 ’50 Smartest Companies’ list, recognizes Deepseek, Game Science and Unitree Robotics
At the EmTech China 2025 Global Technology Summit held last Friday, MIT Technology Review announced its prestigious annual list of the “50 Smartest Companies.” Among the distinguished honorees were Deepseek, Game Science, and Unitree Robotics. Deepseek stood out for its remarkable achievement in algorithm optimization, demonstrating world-class model performance while maintaining low training costs, marking a significant advancement in resource efficiency within the tech industry. The summit highlighted these companies' innovative contributions to technology, underscoring the importance of efficiency and performance in the rapidly evolving landscape.
News Feed
Tesla taps ByteDance, DeepSeek to power AI in China
Tesla, the US automaker, has announced plans to reduce prices for its Model S and Model X vehicles produced in China. However, the company has not disclosed whether it intends to implement similar price adjustments for its other China-made models, such as the Model 3 sedans and Model Y crossovers. This decision comes amid ongoing competition in the electric vehicle market and aims to enhance Tesla's sales performance in the region. The lack of clarity regarding the pricing strategy for its other models leaves stakeholders and potential buyers awaiting further information.
Heavy Hitters News Connected cars Highlight Mobility New Energy Vehicles
DeepSeek credits Tencent for major performance boost in open-source framework DeepEP
Tencent's technology team has enhanced the performance of DeepSeek's open-source DeepEP communication framework, significantly improving its efficiency in various network environments. The Chinese AI startup reported that testing revealed a remarkable 100% performance boost on RoCE networks and a 30% improvement on InfiniBand (IB) networks. These advancements are expected to provide more effective solutions for AI model training. DeepSeek publicly recognized Tencent's contributions to the project on GitHub, highlighting the collaboration's impact on the development of AI technologies.
News Feed
DeepSeek releases new models Janus-Pro and JanusFlow on Lunar New Year’s Eve
In the early hours of January 28, coinciding with Lunar New Year’s Eve, DeepSeek introduced two innovative multimodal frameworks, Janus-Pro and JanusFlow. Janus-Pro serves as an enhanced iteration of the original Janus framework, aiming to provide a cohesive solution for both multimodal understanding and generation. This new model features decoupled visual encoding, which significantly boosts its adaptability and performance across a range of tasks. The launch reflects DeepSeek's commitment to advancing technology in the field of artificial intelligence, particularly in enhancing the capabilities of multimodal systems.
News Feed
Silicon Valley stunned as China’s DeepSeek-R1 surpasses US AI giants, undercuts costs
Tensions escalated in Silicon Valley this week following the launch of DeepSeek's R1 model, a Chinese AI technology that has surpassed leading U.S. companies such as OpenAI, Meta, and Anthropic in third-party performance benchmarks. The unease began earlier with the release of DeepSeek v3, which already outperformed Meta’s Llama 4. The R1 model, unveiled on January 20, has been noted for its significant improvements in model inference capabilities, raising concerns among American tech firms about competition and innovation in the rapidly evolving AI landscape.
News FeedRobotics needs a service framework.
RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.
Join the RobotToday community on LinkedIn
Daily robotics news, in-depth analysis, conference highlights, and discussions with professionals worldwide.
