NAll news

All news

Saturday36 items

  1. Rohan PaulProduct release
    AI score51

    Subvocal launches glasses for silent mouthing input that can transcribe 200+ words per minute

    Subvocal has launched a pair of glasses that can convert silently mouthed words into text at a speed of over 200 words per minute. The principle is that radar passes from the frame through the face, and based on changes in the reflected signal as each word is silently mouthed, AI decodes it into text; at most, bystanders see only slight mouth movements. The product is now available for pre-order, and the team previously raised $1.3 million in funding from YC and Afore Capital.

    Quoted postAarush Yadav · @AarushYadav47

    We’ve raised $1.3M from YC & Afore Capital so you never have to type or speak again. Typing is slow. Talking into your computer is awkward. Subvocal fixes both Our glasses give you the speed of speech without you making a sound or touching a keyboard. It tracks your internal speech muscles with non-contact sensing and decodes silent, almost invisible mouthing into text at more than 200 words/min. Pre-orders are now live at: https://t.co/DzrwIpg3GR

    #Product updates#Speech#English text is being prepared.

Friday141 items

  2. Aravind SrinivasProduct release
    AI score61

    Perplexity Computer Can Control the Local Browser via a Chrome Extension

    Perplexity Computer can now control the user's Google Chrome browser from Perplexity's Mac and Windows apps. After a one-time setup via the new Chrome extension, users can have the agent perform local browsing tasks on Chrome. Perplexity says the feature is now available to all Computer users on the Mac and Windows apps.

    Quoted postComputer · @AskPerplexity

    Computer's Mac and Windows apps now work in your own Chrome. The new Perplexity extension links the two, so Computer uses sites you're already signed into. Here, it turns a business's reviews into a prioritized list of fixes and replies. Live now for all Computer users in the Mac and Windows apps. Get it: https://t.co/IuPEEK6bZ0

    #Product updates#Agents#English text is being prepared.

  3. Rohan PaulProduct release
    AI score58

    BioComputing Launches mischief 1 Video Model API Based on Neuron Experiments

    BioComputing has launched an image-to-video model API called mischief 1, which applies the way human neurons cultured on chips with about 4,000 electrodes process information to the open-source video model Wan 2.2. The model adds less than 0.1% parameters on top of Wan 2.2, and the company says its quality is on par with Wan 2.2 while being 5 times faster, starting at 0.5 cents per second. It is already available on fal and OpenRouter and will launch on AWS Marketplace. The model itself contains no living cells.

    Quoted postAlex Ksendzovsky · @AlexKsendzovsky

    today, we introduce a little mischief to the world. for four years, we’ve been brewing real brain cells on electrodes to understand how they compute. now, we’ve turned those discoveries into a video model anyone can use. from 0.5¢ a second, mischief 1 is at least 5x faster than its base and the cheapest image-to-video model on @fal and @OpenRouter. soon to be on @awscloud marketplace. it’s not a delusion anymore. biocompute is real, commercial and this is just the beginning. happy halloween.

    Same story · 3 reports

    #Model releases#Video#Open-source ecosystem

  4. heise onlineProduct release
    AI score55

    Oracle releases Fusion Companion and Fusion Agentic Home, available…

    At Oracle AI World, Oracle released Fusion Companion and Fusion Agentic Home; employees can describe business goals in natural language, and the software selects and coordinates the required transactions, processes, approvals, and agents accordingly, with entry points including supported AI applications such as ChatGPT. Fusion Companion combines business data, user permissions, and Fusion application context, and preserves existing identity, permission, and approval rules; complex calculations and long-running tasks are handled by the Fusion Claw execution environment announced on September 29, and AI processing uses models such as OpenAI and Google Gemini. Oracle also introduced the Fusion Setup and Configuration Registry under the Universal Permissive License 1.0, requiring Fusion Applications Release 26D and the October 16, 2026 maintenance pack or a newer version; Fusion AI is billed by AI Unit, listed at 1 cent per AI Unit.

    #Product updates#Agents#OpenAI

  5. heise onlineProduct release
    AI score68

    Google releases Android Studio Rabbit 2, opening access to external coding agents such as Claude and Codex

    At the Android Dev Summit, Google released Android Studio Rabbit 2, opening access to external coding agents such as Claude, Codex, Junie, and Antigravity through a Bring Your Own Agent approach, invoking tools such as the emulator, build system, and Logcat via the Agent Context Protocol (ACP) interface. The new version adds Skill Manager to manage Android Skills, covering scenarios such as migrating XML views to Jetpack Compose, upgrading to AGP 9, and adaptive layouts for foldable screens, and provides a /review command to review and fix code written by other agents. The CLI can serve as a plugin for Antigravity, Claude, and Codex, and android preview reduces Token consumption by up to about three times when there is no emulator visualization; Google also demonstrated Android Bench 2, which compares the performance and cost of various models.

    #Product updates#Agents#English text is being prepared.

  6. Unite.AIProduct release
    AI score55

    TBC Launches Neuron-Inspired Video Model Mischief Based on Wan 2.2

    The Biological Computing Co. (TBC) will launch its image-to-video model TBC Mischief on October 30 via fal.ai. The model is based on Wan 2.2, and the company claims it delivers 5x faster generation, 80% lower cost, and better image quality, but has not provided independent benchmarks. Mischief is priced at $0.02 per second of generated video and will be free until November 2; TBC says Wonder Studios generated 100 five-second clips in a batch in about 6 minutes. Its approach uses electrical activity recordings from cultured neural networks to guide research, then modifies an existing diffusion Transformer with a post-training adapter layer that is less than 0.1% of the underlying model's parameter count, with inference still running as software on conventional AI hardware.

    Same story · 3 reports

    #Model releases#Video#English text is being prepared.

  7. MacRumors Apple IntelligenceProduct release
    AI score58

    Draw Things Team Launches Local Code, a Local AI Assistant for Mac

    Draw Things developers have launched Local Code, an open-source AI assistant. Mac users can download models and process prompts and files locally, without uploading to the cloud. The app supports generating presentations, images, and videos, analyzing spreadsheets, handling Word and PowerPoint files, and includes a built-in browser that can browse the web on the user's behalf; with user authorization, it can read information from Mail, Messages, and Notes and modify Calendar and Photos. The app offers Qwen and DeepSeek models; Qwen 3.8 27B requires at least 16GB of memory and a download of about 10GB to 21GB, while the larger DeepSeek models require downloads of about 89GB or 376GB; local use is free, and cloud models can also be used via your own API key or prepaid credits starting at $9.99. The app previously entered public beta on September 12 and is now available as a direct Mac download; the developers say a Mac App Store version is coming soon.

    #Product updates#Open-source ecosystem#English text is being prepared.

  8. UnslothFirst-partyProduct release
    AI score72

    Unsloth releases GGUF version of DeepSeek-V4.1-Flash-1017

    Unsloth releases a new GGUF version of DeepSeek-V4.1-Flash-1017, claiming it ranks highest in benchmark tests. It says Dynamic V3.1 leads on almost all benchmarks including GPQA Diamond, AIME25, LiveCodeBench v6, MMLU-Pro, and Aider, requires more than 64GB of memory to run, and includes links to the GGUF and the blog.

    Quoted postDeepSeek · @deepseek_ai

    🚀 DeepSeek-V4.1-Flash-1017 is here! Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API. 🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model. 1/4

    #Model releases#Open-source ecosystem#DeepSeek

  9. The Beijing NewsProduct release
    AI score58

    Baidu Search upgrades its AI API, with a minimum of 4.9 yuan per thousand calls, lower than DeepSeek

    At its product upgrade briefing on October 30, Baidu announced a fully upgraded Search AI API, with a promotional minimum price of 4.9 yuan per thousand calls, lower than the price of the DeepSeek search API. Baidu says the capability now covers more than 170 tool scenarios, with GUI scenario coverage on the search page increased 5-fold; it connects to over 5 billion resources and data from government affairs, livelihood services, the central bank and others, covers more than 20 industries and opens up its ecosystem. Its cost-effectiveness comes from an AI-native four-layer architecture: the engine layer self-evolves through multi-Agent collaboration, raising overall iteration efficiency by more than 80%, while the application layer supports full access via API, MCP and Skill.

    Same story · 3 reports

    #Product updates#Search#Agents

  10. 36Kr · AIProduct release
    AI score58

    AGILINK Releases OmniGlove Pro Data-Collection Glove; Dexterous Hand Completes 72-Process Production Line Validation

    AGILINK released the OmniGlove Pro data-collection glove and began mass-production delivery at the Dexterous Manipulation sub-forum of the AgiBot Innovation Day, attempting to fill the tactile data gap with a "real-machine Ready" standard. During the same period, its OmniHand dexterous hand continuously completed 72 processes at a POC workstation on the Shaoxing Wolong production line, with a single cycle taking 8.5 minutes, a task success rate of 98%, and the ability to operate continuously for 8 hours. Official data shows that as of October 2026, cumulative mass production of the OmniHand series exceeded 15,000 units, and AGILINK also announced the opening of its tactile data, manipulation models, and industrial deployment framework.

    #Product updates#Embodied AI#English text is being prepared.

  11. gihyo.jpProduct release
    AI score74

    GitHub officially makes Copilot Fusion available at Universe 2026, and opens public beta of Copilot…

    At GitHub Universe 2026, held October 28–29, 2026, GitHub announced that Copilot Fusion, which integrates automatic model selection with multi-model collaboration, is now generally available. It can be used in VS Code, the GitHub Copilot app, the CLI, and JetBrains IDEs, and it automatically determines which model to use and which of three execution modes—Single, Cascade, or Critique—to apply, based on three processing tiers: Standard, High, and Max. The desktop GitHub Copilot app is also beginning a public beta of Projects, which centralizes materials, repositories, sessions, and artifacts in one place for management. GitHub says Fusion Max scores 60 on the Artificial Analysis Coding Agent Index, within 2 points of Claude Code using Opus 5.5 Max, while its average API cost per task is about half and its time is about 40% less.

    Same story · 2 reports

    #Product updates#Agents#English text is being prepared.

  12. English text is being prepared.Product release
    AI score51

    XGIMI Jide Starts Selling AI Memory Glasses G1 and Memory Band L1, Starting at 599 Yuan

    XGIMI Jide's AI Memory Glasses G1 and AI Memory Band L1 officially went on sale on October 28. The Band L1 standard edition is priced at 599 yuan, and the Glasses G1 standard edition has a first-release price of 1,299 yuan. The two devices share the XGIMI Jide App and the same AI memory service, recording as soon as they are powered on, and can transcribe daily conversations and automatically generate meeting minutes, to-do items, and life cards. For data management, it adopts an approach of keeping only text and not the original audio; after voice transcription is completed, the original audio is deleted, and it supports setting recording-free periods and privacy-blocked words.

    #Product updates#English text is being prepared.#Speech

  13. Intelligent Manufacturing NetworkProduct release
    AI score52

    Daxiao Robotics and Nanyang Technological University Jointly Release World Model Mundo

    Daxiao Robotics, in collaboration with the S-Lab at Nanyang Technological University, has released Mundo, a world model built for specific environments that supports scene reconstruction, free-viewpoint exploration, and spatiotemporal simulation. Mundo can write physical properties such as object mass and friction into scene programs, with a physics engine computing motion, contact, and collision, and supports invoking agents to edit objects, properties, and actions in the scene using code. Inputting 1 to 12 images is sufficient to construct an explorable three-dimensional space; the same set of reference images supports multiple camera paths, and larger scenes can continuously generate exploration footage ranging from tens of seconds to the minute level.

    Same story · 3 reports

    #Model releases#Embodied AI#Multimodal

  14. China.com TechnologyProduct release
    AI score60

    Tencent Beacon AI Officially Opens to the Public: Complete Data Analysis with a Single Sentence

    Tencent announced that its AI data analysis platform Tencent Beacon is officially open to the public. Enterprise and individual users can register and use it via beacon.qq.com. The platform offers capabilities such as natural language data querying, dashboard generation, attribution analysis, metric monitoring, and team collaboration. Beacon AI supports integration with 10 data sources across 3 major categories. According to the company, building a full-chain operations dashboard now takes no more than 2 hours, down from 4 to 6 working days with traditional BI, and a business monitoring dashboard takes 8 minutes from opening the dialog box to finalization. The platform comes with 12 preset data analysis skills and 3 visualization skills, and can connect to enterprises' internal systems via the MCP protocol. New users receive 500 credits upon first registration, and 500 credits per month thereafter.

    #Product updates#English text is being prepared.#English text is being prepared.

  15. Sina TechProduct release
    AI score65

    Zhiyuan Unveils GO-3 Preview Foundation Model, Genie Master Framework, and Genie G2 Max Robot

    At the Embodied Work Productivity sub-forum of the 2026 Zhiyuan Innovation Day, Zhiyuan unveiled the embodied foundation model GO-3 Preview, the embodied Agent framework Genie Master, and the heavy-load embodied industrial robot Genie G2 Max. GO-3 Preview adopts the unified WRAM architecture for understanding, prediction, and action generation; on 100 tasks across 20 skill categories, its Zero-Shot average success rate rose from 13.4% to 56.4%, and in model-scale expansion experiments it rose from 56.2% to 73.2%; it has also topped the three major evaluation leaderboards RoboDojo, RoboTwin, and RoboColiseum. Co-President Yao Maoqing proposed three capability standards—measurable, replicable, and sustainable—and the event also presented data including a 98.3% success rate on a phone packaging task and the Genie G2's convenience-store task success rate rising from 28.6% on the first day to 85.7% by the fifth day.

    Same story · 12 reports

    #Industry news#Embodied AI#Agents

  16. LeiphoneProduct release
    AI score67

    NVIDIA and MIT Team Open-Source a 4K Video Enhancement Solution Based on MiniMax H3

    The Efficient-Large-Model team, composed of researchers from NVIDIA and the Massachusetts Institute of Technology, has released the model weights for the video enhancement project Open-H3-Regenerate-2K on Hugging Face, and the inference code has been made available in the NVlabs/Sana repository. The solution is based on MiniMax H3 and supports 2560×1440 2K output, and can achieve native 4K regeneration at up to 3840×2160. During training, the H3 Transformer backbone and the video VAE are frozen, and only the newly added refinement LoRA is trained. Team tests show that on a single NVIDIA B200 GPU, the combined time for upscaling, three-step refinement, and 2K decoding of an approximately 5-second video dropped from 87.26 seconds to 22.45 seconds; on RefinerBench-150, the solution's UniPercept average score is 60.0, compared with 53.4 for the official H3-Regenerate-2K API.

    Same story · 2 reports

    #English text is being prepared.#Video#Open-source ecosystem

  17. LeiphoneProduct release
    AI score60

    Xingyuan Intelligence Releases Cross-Embodiment Data Engine Chorus, Untangling the Challenges of the Embodied Data Pipeline

    Xingyuan Intelligence has released Chorus (Zhongxing), a cross-embodiment data engine for robot pre-training, which can convert human manipulation data collected by devices such as smart glasses, chest-mounted cameras, wrist cameras, and UMI into training data adapted to different robot embodiments. Xingyuan claims that the average usable data conversion rate has risen to 85%, that it adapts to 16 robot embodiments under the current test configuration, and that the processing cost per hour of effective robot trajectories has dropped to 15 yuan; combined with 3x instruction augmentation, this can achieve roughly 40x expansion of effective training data. The article also outlines the difficulties in the five stages of embodied data—filtering, annotation, motion conversion, validation, and feedback iteration—as well as the respective shortcomings of four types of data sources: real robots, simulation, UMI, and human first-person videos.

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

  18. OpenBMBProduct release
    AI score58

    OpenBMB Reposts Real-World Test Case of MiniCPM-OCR on Handwritten Notes

    OpenBMB reposted a user's real-world test case of using MiniCPM-OCR to recognize actual handwritten notes, and thanked them for sharing. The quoted tweet explains that MiniCPM-OCR consists of one model responsible for reading documents and another model that evaluates the output and flags parts that may be wrong, and is open-sourced under the Apache 2.0 license.

    Quoted postHasan Toor · @hasantoxr

    OpenBMB just released an OCR parser alongside a separate model that evaluates its outputs. It’s called MiniCPM-OCR. One model reads your documents. A second model tells you which parts it probably got wrong. 100% open source. Apache 2.0. ↓ https://t.co/ihbtDjZaUy

    #Industry news#Multimodal#Product updates

  19. QbitAIProduct release
    AI score57

    NVIDIA and MIT teams' Open-H3-Regenerate-2K, an open-source video enhancement solution based on MiniMax H3…

    The Efficient-Large-Model team, composed of researchers from NVIDIA and MIT, has released the model weights for the video enhancement project Open-H3-Regenerate-2K on Hugging Face, with the accompanying inference code now available in the NVlabs/Sana repository. Based on MiniMax H3, the solution regenerates low-resolution video drafts into high-resolution video, supports 2K output at 2560×1440, and can achieve native 4K regeneration at up to 3840×2160. The team kept the H3 Transformer backbone and video VAE parameters unchanged, training only the newly added refinement LoRA; at inference time it is merged with community adapters such as LightX2V FL2VA Turbo and LMS Sharp, regenerating image details through three-step denoising. On a single NVIDIA B200 GPU, the combined time for latent-space upscaling, three-step refinement, and 2K decoding of an approximately 5-second video dropped from 87.26 seconds to 22.45 seconds, about 3.9 times faster than before; on RefinerBench-150, UniPercept averaged 60.0, compared with 53.4 for the official H3-Regenerate-2K API.

    Same story · 2 reports

    #English text is being prepared.#Video#Open-source ecosystem