deepseek

DeepSeek announced the full open-source of its internal inference framework DeepInfer, and simultaneously open-sourced the production-level token encoding and decoding component deepseek-recipe. DeepInfer integrates high-performance operators and a deeply optimized inference engine, supporting distributed inference, speculative decoding, SWA Bounded Replay, and structured output, compatible with NVIDIA and Huawei Ascend platforms.DeepSeek stated that it has implemented high throughput and low latency large-scale inference services for DeepSeek V4.1 Flash based on DeepInfer, and future inference code and optimizations for new models will continue to be updated here. deepseek-recipe is fully aligned with the processing logic of the DeepSeek official API, aiming to ensure that the community receives an experience consistent with the official deployment. The open-source repository and technical report have been released simultaneously.

According to Beating reports, Tencent announced the opening of Yuanbao Search to developers through Tencent Cloud, supporting API, MCP, and Skill access. Tencent disclosed that this search service has an index of hundreds of billions, with an average daily call volume of hundreds of millions, and has supported over 50 internal businesses including WorkBuddy, ima, and Tencent Video. The underlying technology of Yuanbao Search comes from Sogou Search, integrates Tencent's content ecosystem, and connects to 157 authoritative sources across 13 industries including finance, healthcare, and law.Records from the Tencent Cloud Internet Search API show that this service went live on July 1, 2025. Alibaba Cloud launched its independent Internet Search API on March 20 of the same year; Volcano Engine upgraded its existing service to Doubao Search in June this year and officially released it to the public in July; DeepSeek opened its self-developed search engine's Search and Fetch API on October 17. The prices for Yuanbao Search's Express, Standard, and Premium versions are 14 yuan, 20 yuan, and 26 yuan per thousand calls, respectively, with an additional monthly free quota of 2,000 calls; the Express version only supports public web sources. The price for DeepSeek Search API is 5 yuan per thousand successful requests, Alibaba Cloud OpenSearch Lite is 12 yuan, and the Pro version was reduced to 18 yuan on October 15.

The AI model evaluation platform Arena announced that DeepSeek-V4.1-Flash-20261017 (Max) has ranked first among open-source models in Agent Arena and 11th overall. Based on approximately 7,500 real usage sessions, this model shows a comprehensive net improvement of about 5.1% compared to the average model, with a median single-task cost of about $0.20.This version ranks first among open-source models in four metrics: task success confirmation, user positive and negative feedback, reliability of Bash command execution, and guideability, with the task success confirmation metric ranking 8th overall.

The spokesperson of the Ministry of Foreign Affairs, Mao Ning, hosted a regular press conference. A reporter from AFP asked whether the U.S. government has requested federal agencies to stop using Chinese artificial intelligence products like DeepSeek, citing concerns about the potential transfer of sensitive data, and inquired about China's comments.Mao Ning stated that China has always opposed the generalization of the concept of national security and the politicization of technological and economic issues, and will firmly safeguard the legitimate rights and interests of Chinese enterprises. Previously, the U.S. government prohibited civilian agencies from using DeepSeek products.

According to The Information, DeepSeek is collaborating closely with Huawei to jointly design the next generation of AI chips, aiming to build models that do not rely on NVIDIA's computing power. Insiders revealed that Huawei engineers have entered DeepSeek's office to understand its model architecture and algorithm requirements, and to test how the chips handle data transmission and compression, with the first compatible chip targeting Ascend 950DT.The report pointed out that this move comes as China pushes to reduce its dependence on American chips. Limited by TSMC's foundry services and U.S. export controls, Huawei's Ascend 950DT is planned for release in 2027, but mass production may be delayed until 2028. To address the chip shortage, DeepSeek is using the TileLang programming language to bypass CUDA restrictions, while still attempting to acquire the banned NVIDIA Blackwell chips through third-party data centers. In addition, DeepSeek is forming a chip team and has begun testing clusters with over 1,000 processors.

The U.S. Department of Homeland Security (DHS) instructed U.S. civil agencies on Monday to remove products from the Chinese AI company DeepSeek from government IT systems, as well as any third-party services built on that model, and to prohibit new procurements. The notice stated that using DeepSeek could lead to the transfer of sensitive data to China and could generate intelligence targeting U.S. individuals and institutions. DHS indicated that these models constitute "multiple threat vectors."The ban targets not only DeepSeek's own applications but also covers third-party services based on its development. A DHS memorandum from July was made public on Monday, which had previously requested agencies to monitor the risks related to DeepSeek in the traffic in and out of government websites. The U.S. Department of Commerce, the Navy, and the states of New York, Tennessee, and Virginia had previously restricted DeepSeek. The Federal Communications Commission (FCC) will vote on October 29 to decide whether to ban all Chinese laboratories from testing electronic products used in the U.S.DeepSeek is a startup based in Hangzhou, and its open-source model attracted widespread attention on January 27, 2025, with development costs of less than $5 million, leading to a 16% drop in Nvidia's stock price that day, evaporating about $600 billion in market value. In April 2025, the U.S. House of Representatives' Special Committee on China Issues stated that DeepSeek might secretly send U.S. user data to the Chinese government and may have used about 60,000 Nvidia chips, of which 20,000 were subject to U.S. export controls. The V4-Pro model preview released by DeepSeek in April this year was priced about 98% lower than OpenAI's GPT-5.5 Pro and was made available for free download and modification under the MIT license; this week it also released an updated version of the V4.1-Flash model focused on efficiency optimization.

According to Beating reports, Lei Feng Network interviewed several headhunters and found that major domestic internet companies are tightening their recruitment quotas but are still willing to break salary limits for top AI talents. Headhunters revealed that this year, some outstanding fresh graduates in foundational model directions have annual packages reaching 2 million yuan; ByteDance Seed and Xiaohongshu's highest offers have reached 4 million to 5 million yuan, while Baidu is in the range of 2 million to 3 million yuan.The high salaries for talent are exacerbating internal salary discrepancies. An internet professional stated that some employees in foundational models have lower job levels but earn more than their higher-level colleagues. Similar situations emerged as early as 2017; an experienced engineer recalled that at that time, a returnee fresh graduate who joined a year later and was at a lower job level had a monthly salary 10,000 yuan higher than his own, leading to a monthly turnover of about 10% in a team of over thirty people.Headhunters noted that application-side positions are shrinking, while foundational model teams still lack individuals who can truly train large models. Coding Agents, model evaluation, pre-training data, and Agent infrastructure have become new focal points for talent acquisition; DeepSeek has recently been expanding its recruitment for backend developers and Agent Infra engineers, while the hiring enthusiasm in areas like AI search has decreased. Companies are not only looking at academic qualifications and papers but also place greater importance on whether candidates have actually trained large-scale models in leading teams; the salaries and responsibilities of scarce talents can be negotiated separately, while competition among ordinary AI job seekers is more intense.

On October 19, the B.AI platform announced that DeepSeek V4.1 Flash has seamlessly upgraded to the latest V4.1 Flash-1017 version, further optimizing response speed, inference quality, and operational stability, bringing a more efficient and intelligent AI experience to developers and users.This upgrade continues the seamless switching method, allowing users to keep the same model ID, and existing API calls and workflow configurations can remain unchanged, enabling direct access to the new model capabilities, significantly reducing migration and adaptation costs. B.AI stated that it will continue to follow the iteration of cutting-edge models, providing developers with more stable and cost-effective model services.

According to Bloomberg, Asian semiconductor stocks continued to weaken on Monday. The Bloomberg Asia Chip Index fell more than 2%, reaching its lowest level since July, with TSMC, SK Hynix, and Advantest being the main drags. The Korea Composite Stock Price Index recorded its largest weekly decline since October 2008 last week, dropping more than 3% at one point on Monday before recovering some ground; the Nikkei 225 index fell 2.3% at one point, and the Taiwan Weighted Index dropped 2%.Market concerns stem from two aspects. First, the pressure on financing for AI infrastructure in the United States continues to ferment, with CoreWeave's financing capabilities being questioned and the IPO expectations for Anthropic and Nscale cooling. Second, Chinese AI company DeepSeek released a new model update over the weekend, claiming faster operation speeds and lower resource consumption, raising concerns about the potential decrease in storage capacity required for AI models and the possible impact on chip demand.Kang DaeKwun, CEO of Life Asset Management, stated that the market uncertainty is too high, making it difficult to determine when the Korean stock market will bottom out. Financing for U.S. corporate AI projects is one of the main risks, and the profitability of Korean chip companies still heavily relies on continuous investment in AI infrastructure by U.S. companies. This week, the earnings reports from tech giants like Meta and Tesla will be an important window to observe the growth momentum of AI capital expenditures.

According to a report by Jinshi, Citic Securities' research report indicates that Harness is expected to become a key component for DeepSeek in transitioning from model supply to platform ecology, transforming open-source and engineering advantages into user base, ecological influence, and commercialization capabilities. As competition among Agents expands from model capabilities to stable and user-friendly task delivery, DeepSeek is expanding user and task coverage through plug-ins, scenario presets, and cross-Agent collaboration, extending its infrastructure advantages from hardware-software synergy to complete task execution.The research report states that its long-term value lies in the accumulation of demand through ecological expansion, enhancing product stickiness through configuration and process sedimentation, and supporting safe governance with traceability capabilities, collectively creating conditions for sustained use and enterprise deployment. Based on official model calls, the company can also explore platform services and ecological transactions, transforming ecological accumulation into more diverse revenue sources and supporting reinvestment in products and ecology. Citic Securities is optimistic about the continued rise of domestic large models.

DeepSeek has announced that the DeepSeek API now supports web search and crawling, charging based on the number of requests, with a unit price of 5 yuan for every 1000 successful requests.The announcement states that current users accessing the Search/Fetch interface through the Server-side tools in DeepSeek Harness/Claude Code/Codex will not be charged for these Search/Fetch interface calls at this time, and fees will start to be charged per call from 00:00 Beijing time on November 1, 2026.

On October 16, B.AI announced the domestic model brand Token weekly ranking from October 9 to 15. Xiaomi model Token usage has remained steady at 2nd place for three consecutive weeks, with a month-on-month increase of 49.9%, and the platform Token usage proportion increased by 10.4 percentage points; DeepSeek continues to lead this week with a total of 1.03T Tokens; Zhipu saw a month-on-month growth of 44.3%, with the platform proportion increasing by 3.7 percentage points, ranking 3rd; Moonshot Token usage surged by 286.5% month-on-month. With multiple models competing on the same stage, real usage witnesses market changes. B.AI continues to expand model choices, helping developers flexibly access diverse AI capabilities. Welcome to log in to b.ai to explore.

The Tencent Cloud WorkBuddy and CodeBuddy teams released a notice regarding "Explanation of DeepSeek Model Service Abnormalities and Compensation Announcement." In response to the previous abnormalities in the DeepSeek model service, the team will fully refund the points consumed by affected users during the period from October 8 to 14.The announcement also offers an additional 500 points to all users of the domestic versions of WorkBuddy and CodeBuddy. Previously, DeepSeek apologized for the service system failure and announced a compensation plan, stating that the consumption amount of affected user accounts within 7 days (from October 8 to 14) will be fully refunded in the form of bonus points.

DeepSeek has integrated the Harness team along with the Web, App, and Platform teams into the Product Department, responsible for all product, design, and development work of the company, led by Cui Tianyi as the head of the Product Department.DeepSeek is currently hiring product managers, product designers, and project managers, with product managers divided into two directions: AI products and Agent Harness, covering web, App, open platform, and Harness product lines. Job requirements include experience in complete delivery, experience with Agent-type products, as well as skills in prototyping, data analysis, and communication within open source communities. The work locations are Beijing and Hangzhou, with both full-time and internship positions available.

According to Top Chinese Innovation Society reports, former head of the basic model for humanoid robots at ZhiJi Dynamics, Yu Tao, has left to start his own venture, focusing on the combination of LLM and embodied intelligence. According to relevant sources, his company's first round valuation is 70 million USD, and the second round valuation is 200 million USD, with DeepSeek and a leading top-tier fund participating in the investment.Yu Tao holds a PhD from the University of Science and Technology of China and was a visiting scholar at the National University of Singapore. He previously served as a senior researcher in Tencent AI Lab's Absolute Enlightenment team, where he led the team to build the full-body tracking system LimX WBT and the motion generation system from scratch during his time at ZhiJi Dynamics. According to reports from Phoenix Technology, DeepSeek founder Liang Wenfeng has recently met with several entrepreneurs in embodied intelligence and is internally evaluating different entry methods, such as forming a team independently or investing in external startups.In August of this year, DeepSeek invested approximately 141 million RMB through a strategic placement in Yushu Technology for its IPO; previously, Bloomberg also disclosed its plan to lead the investment in the robotics company Winverse AI founded by former Tencent AI scientist Hu Han.

DeepSeek API Platform has released a notification regarding service recovery and compensation, stating that between 8:11 and 13:04 Beijing time on October 15, 2026, its service system experienced a failure, resulting in the API service being unavailable during that period. The platform apologizes for this and has stated that a comprehensive investigation and rectification regarding service stability has been initiated, with efforts to streamline and eliminate single point dependencies to prevent similar issues from occurring again.As compensation, the platform will fully refund the consumption amount of user accounts within 7 days (from October 8 to October 14) in the form of bonus credits, which will be automatically credited to the accounts. API billing deductions will first be taken from the bonus account.

OpenCode announces a limited-time free access to DeepSeek V4.1 Flash. Users can select DeepSeek V4.1 Flash Free from the model list, with no charges for input and output tokens. The free version has a context limit of 200,000 tokens, while the paid version has a limit of 1,000,000 tokens, which is an 80% reduction, making it easier to encounter context limits when processing large code repositories or running a Coding Agent for extended periods.DeepSeek V4.1 Flash was integrated into OpenCode on September 10. On October 12, OpenCode will increase the usage quota for V4.1 Flash in the $10 Go package to a value of $80, and subsequently add an independent free tier. The official documentation confirms that this free access has a time limit, but the end date has not yet been announced, and states that the free tier service provider adheres to a zero data retention policy and will not use user content for model training.

DeepSeek's website and API services experienced partial interruptions starting from October 15 at 08:11, affecting the DeepSeek V4 Pro API, V4.1 Flash API, and conversation services. Services have now been restored, and the interruption lasted a total of 5 hours.

According to the official status page of DeepSeek, the DeepSeek website and API services experienced partial interruptions starting from October 15 at 08:11, affecting three components: DeepSeek V4 Pro API, V4.1 Flash API, and the conversation service.As of 09:08, the official status update is "Identified," indicating that the cause has been determined and repairs are being implemented. Previously, from 08:11 to 08:14, the status was displayed as "Under Investigation." Currently, the service remains in a "Complete Outage" state, and the recovery time has not been announced.