NAll news

All news

Tuesday109 items

  5. Zhidongxi
    AI score78

    DeepSeek-V4.1-Flash-1017 Tops Code Arena WebDev Open-Source Leaderboard

    Arena announced that DeepSeek-V4.1-Flash-1017 (Max) surpassed open-source models such as Kimi K3 (Max) with a score of 1670, becoming the highest-ranked open-source model on the Code Arena: WebDev leaderboard, and also surpassed closed-source models such as Claude Opus 5 and Grok 4.7, trailing GPT-6 Sol by only 12 points. The model was launched on October 17, priced at 2 yuan per million tokens for input (cache miss) and 8 yuan for output during peak hours, and 1 yuan for input and 4 yuan for output during idle hours, with a comprehensive cost of about 1/10 that of Kimi K3. Zhidongxi used it to create two cases, an ancient Kaifeng 3D scene and a web-based racing game, both completed in a single attempt, at a total cost of about 6 yuan.

    Same story · 2 reports

    #Evaluations / benchmarks#English text is being prepared.#Open-source ecosystem

  12. Caixin
    AI score71

    US Department of Homeland Security bans federal civilian agencies from using DeepSeek and related products

    The US Department of Homeland Security publicly issued a ban on October 19, requiring US federal civilian executive agencies to stop procuring DeepSeek-related products and services and to remove DeepSeek-related products from existing information technology systems, including third-party products and services developed using DeepSeek solutions. According to the US Cybersecurity and Infrastructure Security Agency list, federal civilian executive agencies include 102 agencies such as the Department of the Treasury, the Department of Justice, the Department of State, and the Department of Agriculture. DeepSeek had previously been banned by the US Department of Commerce, congressional offices, the Pentagon, NASA, the US Navy and other departments; this time the ban has been expanded to all US federal civilian agencies.

    Same story · 15 reports

    #English text is being prepared.#DeepSeek

  13. Tech.eu
    AI score54

    Former Google DeepMind researcher founds London AI company Metis Reasoning, focusing on compute-efficient reasoning

    Metis Reasoning, a London AI company co-founded by former Google DeepMind research lead and AlphaGo co-creator Thore Graepel, has officially launched, seeking a compute-efficient path distinct from the frontier models of Anthropic and OpenAI, and is believed to have raised tens of millions of dollars in initial funding. The company has a team of 12 people, including researchers, mathematicians, robotics experts, and technology entrepreneurs. It believes that LLMs have fundamental limitations in reasoning, that results are difficult to verify, and that there is a lack of audit trails; its approach is more verifiable, more robust, and safer, and it says it has already achieved some positive results on frontier reasoning problems at a cost far below that of frontier models. Metis has not yet announced a release date for its first AI model.

    #Industry news#Reasoning#Agents

  15. Unite.AI
    AI score52

    Town Completes $108 Million Series B Funding, Valued at $1 Billion

    Town announced on October 19, 2026 that it had completed a $108 million Series B funding round, co-led by Forerunner Ventures and Menlo Ventures, at a valuation of $1 billion, with participation from existing investors including a16z and First Round Capital. Town's AI work assistant Townie can now collaborate directly with one another, asking each other questions, connecting context across teams, and handing off work, while still sharing information only within the scope authorized by each respective user. The company says that since July 2026, its paid customers have grown 4x and paid teams 9x, and it supports more than 60 integrations.

    #Industry news#Agents

  16. Unite.AI
    AI score65

    Meta Details Prometheus Supercomputer: Has Trained Muse Models, Plans to Scale to About 1 Million GPUs

    Meta's engineering team has published technical details of the Prometheus supercomputer, confirming that the first phase has trained the Muse family of foundation models (including Muse Wonder) and that it plans to scale the system to about 1 million GPUs. The cluster is located in New Albany, Ohio, currently consists of hundreds of thousands of GPUs, and runs as a single machine; Muse Wonder achieved pretraining goodput above 85% and reinforcement learning above 90%. Meta also disclosed two platforms—Catalina based on NVIDIA GB200 NVL72 and Clemente based on GB300 NVL72—as well as the upcoming Vera Rubin NVL72 platform.

    #Industry news#English text is being prepared.#English text is being prepared.

  17. Unite.AI
    AI score57

    Tenstorrent Completes $150 Million Series E Funding at a $6.2 Billion Valuation

    Tenstorrent announced the completion of a $1.5 billion Series E funding round at a post-money valuation of $6.2 billion, led by Trace Capital and Innovation X. The funds will be used for large-scale deployment of Galaxy servers, continued support for open-weight models for large language models and video generation, and advancing externally licensed AI and CPU intellectual property. The company builds AI processors and systems based on RISC-V, and licenses the TT-Ascalon RISC-V CPU and Tensix AI cores to chip designers such as Samsung and LG.

    Same story · 2 reports

    #Industry news#English text is being prepared.#English text is being prepared.