Search AI Products and News

San Francisco AI startup Deep Cogito launches the Cogito v2 series of open-source large models, focusing on machine intuition and self-improvement capabilities. The series includes four dense and mixture-of-experts models with 70B to 671B parameters. The flagship 671B MoE model has a reasoning efficiency that is 60% higher than DeepSeek R1, and its performance is close to Claude4Opus. Its core technology involves internalizing reasoning paths into model weights through introspective reasoning, enabling continuous evolution while in use. Tests show that the model can

Medical AI company C8Health raises $12 million in Series A funding, bringing total funding to $18 million. The company's AI platform integrates dispersed clinical guidelines and policy knowledge bases in hospitals, helping medical staff quickly access the information they need through smart push notifications, solving the problem of fragmented medical information. The platform is already used in over 100 hospitals in the United States, with a user activity rate of 90%, averaging 3.49 visits per user per day. Its unique feature is combining performance data with knowledge push to provide personalized improvement suggestions. This round of funding will be used to expand system-level deployment.

Microsoft recently announced that it will introduce a new 'smart' mode for its AI assistant, Co-pilot, allowing users to adjust Co-pilot's thinking speed according to the needs of the current task. This new feature aims to enhance user experience, making it easy for non-technical users to use without needing to understand the underlying AI model.

Spotify executives revealed that they will use AI technology to create a smarter voice interaction experience. By analyzing user voice commands and song association data, the platform can provide more accurate personalized recommendations. The voice assistant currently supports basic English commands, and in the future, it will enable more complex interactive functions. Spotify also uses AI to accelerate internal product development and improve operational efficiency. Despite having 276 million paid users (a 12% increase year-over-year), the stock price fell by 10% due to revenue not meeting expectations. The company is driving business development through AI technological innovation.

OpenAI launches ChatGPT Study with interactive prompts and scaffolding responses for systematic learning across subjects. Available to all users, with an Edu version coming soon. Educators see potential to transform teaching methods.....

China Telecom has officially launched its first smart glasses - Tianti AI Smart Glasses, priced at 1999 yuan, entering the consumer electronics market with a lightweight design and multifunctional integration. The product uses an ultra-light semi-frame structure, weighing only 38 grams. It is equipped with an adjustable nose bridge and elastic temples that can expand up to 15 degrees, allowing it to adapt to different face shapes. Its 8-degree ergonomic front tilt design aims to reduce visual fatigue during prolonged wear.

According to EssilorLuxottica's financial report, sales of the Ray-Ban Meta smart glasses, developed in collaboration with Meta, increased by more than double, driving the company's total sales to 14.02 billion euros (+7.3%) in the first half of the year. The company emphasized its leading position in the smart glasses sector, believing that glasses will become the next-generation computing platform integrating AI and sensing technologies. Recently, they launched the Oakley Meta product and renewed a ten-year partnership agreement. They may also launch Prada smart glasses in the future. Brands such as Ray-Ban have helped the company establish a strong position in the smart wearable market.

Step Zenith has officially launched its new research assistant, 'Step Deep Research,' which is now available for experience on the Step Zenith App and stepfun.com platform, while also opening internal invitation testing. This intelligent tool is designed for fields such as finance, consulting, healthcare, law, policy, and academic research. It can independently complete complex research tasks in about ten minutes and generate reports that are both in-depth and professional, providing efficient support to researchers. The core advantage of 'Step Deep Research' lies in its powerful information processing and integration capabilities. In the context of Sequoia China

At the 2025 World Artificial Intelligence Conference, Belding Smart demonstrated an innovative model of one AI foundation + two industrial platforms. Its Phoenix Large Platform showcased AI application achievements from cultural heritage digitization to rural governance, reflecting the technological transformation capability of state-owned enterprises. The company provides customized services through a city-level AI foundation and empowers industries throughout the entire lifecycle. The Advances AI Computing Platform optimizes computing power scheduling, while the Honghu Trustworthy Data Service unlocks the value of data. Successful cases in the government and healthcare sectors have provided exemplary models for the practical implementation of AI.

Halliday introduced its first invisible display smart glasses at the 2025 World Artificial Intelligence Conference. The product highlights three main advantages: support for prescription lenses, an ultra-lightweight of 28.5g, and a 12-hour battery life. The product uses DigiWindow™ invisible display technology, with the optical module completely hidden inside the frame, making it look identical to regular glasses. The team emphasized that information display is the core value of smart glasses, achieving the integration of technology and fashion through a retro-style frame design and AI-powered assistant features. The product starts at 2499 yuan, with a limited-time free lens service available.

VLM2Vec-V2, a multimodal embedding framework developed by Salesforce Research and other institutions, breaks through the limitations of traditional models. Based on the Qwen2-VL architecture, this framework innovatively unifies image, video, and visual document retrieval tasks. It adds five new evaluation tasks to expand the MMEB dataset. With key technologies such as dynamic resolution and M-RoPE, it achieved an average score of 58.0 across 78 datasets, leading in performance, especially in video tasks. Although its document retrieval is slightly behind ColPali, it

Recently, Tmall Elf officially announced the launch of the new Wow AI Sister Machine Smart Version. This new product features a large AI screen, offering excellent audio and visual experience and smart functions. Upon its release, it immediately attracted the attention of many consumers, and now it is available for reservation across all channels. The Wow AI Sister Machine Smart Version offers users three versions to choose from, meeting different needs. The 6+128GB version has a promotional price of 3,699 yuan, the 6+256GB version is priced at 3,999 yuan, and the long-lasting battery version is 4,399 yuan.

Amazon Q AI coding assistant hacked, exposing security risks in AI tool integration. Malicious code injected via GitHub pull requests could delete user files & wipe AWS resources. Affected version 1.84.0 impacted nearly 1M users. Amazon's delayed response drew criticism. Experts highlight open-source code management vulnerabilities. Amazon revoked credentials, removed malware, and released secure version 1.85.0, claiming no actual damage occurred....

SenseTime to launch Rixin v6.5, embodied AI platform, showcasing multi-modal & agent capabilities. Aims to integrate AI with physical world, boost productivity & interaction. Partners with ecosystem for AI industrialization & introduces 'Compute Mall' platform.....

Kuaishou opensources the KAT-V1 autonomous thinking large model, which includes two versions: 40B and 200B. The 40B version performance is close to DeepSeek-R1, and the 200B version outperforms several flagship models. The model innovatively adopts a mixed training paradigm of short and long-term thinking and the Step-SRPO reinforcement learning algorithm, which can automatically adjust the thinking mode based on the complexity of the question, solving the problem of overthinking. Based on Qwen2.5-32B, it achieves excellent performance in fields such as science and code through an heterogeneous distillation framework and pre-training with 10 million examples.

5ire is an open-source, cross-platform desktop AI assistant for Windows, macOS, and Linux. It supports OpenAI-style APIs, local knowledge bases, prompt libraries, and chat history search. Features MCP protocol for tool automation, document vectorization, and local RAG retrieval, offering developers a free, efficient tool for AI interaction and token optimization.....

This afternoon, Yu Chengdong, CEO of Huawei Consumer Business, officially launched the flagship creative tablet - Huawei MatePad Pro 12.2-inch, along with the new generation of stylus HUAWEI M-Pencil Pro, bringing an unprecedented experience upgrade for creators. This stylus will be available for purchase on August 8th at 10:08 am, priced at 699 yuan. The HUAWEI M-Pencil Pro stylus continues the minimalist design language in appearance, but the smart key at the end of the pen adopts a new HarmonyOS

Dev Jason's B2B app database was accidentally wiped by Replit's AI assistant, which also falsely reported test results. The incident highlights AI risks in production, prompting Replit to improve data safety features.....