Search AI Products and News

Kuaishou and Shanghai Jiao Tong University announced the open-source multimodal generation model Orthus at the ICML conference. The model is based on an autoregressive Transformer architecture and demonstrates excellent performance in text-image conversion tasks, with significantly improved computational efficiency compared to existing models. Orthus uses a unique architectural design, separating processing of text and image features through specific modal generation heads, and efficiently fusing them in a unified representation space. Research shows that the model outperforms professional models in image understanding and text-to-image tasks, while also demonstrating strong potential in application scenarios such as image editing, making it a breakthrough in multimodal generation.

On July 23, Kunlun Group introduced two innovative AI models to global users - the Mureka V7 Music Model and the Mureka TTS V1 Audio Model. The launch of these two models marks an important step forward in the field of AI music generation. Since its release, the Mureka V7 Music Model has attracted widespread attention, gaining extensive praise for its efficient music creation capabilities and outstanding music quality. Users simply need to input lyrics, select a style or theme, and click generate to obtain a complete and pleasant song in a short period of time.

The夸克 Health Large Model has passed the written evaluation of 12 core medical disciplines by chief physicians in China, becoming the first large model to reach this level domestically. The model has been integrated into夸克AI Search, allowing users to access chief physician-level AI medical services through deep search. Its core breakthrough lies in establishing a slow-thinking capability, processing complex medical issues through chain reasoning and clinical deduction path modeling.夸克 adopts a dual data production line + dual reward mechanism training system, with a team of thousands of physicians labeling the data, including over 400 deputy chief physicians and above. The platform currently has more than 20 million monthly active medical student users.

The Meta research team introduced an innovative architecture called AU-Net, breaking through the limitations of traditional tokenization techniques. The model learns directly from bytes, dynamically combining bytes into hierarchical sequence representations through a self-regressive U-Net structure. Its contracting path compresses bytes into semantic units, while the expanding path restores information and fuses details, with skip connections preserving local features. A multi-stage pooling strategy is used to extract semantics at the word and phrase levels, and multi-linear upsampling optimizes information fusion. The self-regressive generation mechanism ensures text coherence and significantly improves inference efficiency. This architecture

The Kimi team released the Kimi K2 technical report. This trillion-parameter large model, with 32 billion activated parameters, won the global open-source model competition, with performance comparable to top closed-source models such as GPT4.5. Its innovations include: using the MuonClip optimizer to process 15.5 trillion token data without loss; developing a multi-domain tool data synthesis pipeline; pioneering the restatement method to improve training efficiency, achieving the effect of ten rounds of original data in one round of rewritten data; optimizing model performance through supervised fine-tuning and reinforcement learning. The model is based on N

Latent Labs launches a revolutionary AI model called LatentX, focusing on innovative protein design. The model, developed under the leadership of Simon Kohl, a former DeepMind core member, can generate new protein structures with atomic-level accuracy and high experimental feasibility. Unlike prediction tools like AlphaFold, LatentX has the ability to creatively design proteins. The company uses a licensing model, offering services to academic institutions and pharmaceutical companies, supported by $50 million in funding. The future plans include launching premium paid features.

Alibaba's Qwen3-Coder, a 48B-parameter MoE-based AI model, excels in complex programming tasks with 256K context support. Trained on 7.5TB data, it outperforms GPT4.1 in benchmarks and is open-sourced on ModelScope and Alibaba Cloud.....

On July 22, Tencent HuanYuan announced that its self-developed ASR (Speech Recognition) large model has been officially applied to the ima platform, achieving voice input functionality on mobile app interfaces for the first time, offering users a convenient experience of 'speaking fluently'. Users can now directly input questions or record inspirations through voice, without manually typing on the keyboard, greatly improving input efficiency. Tencent HuanYuan's ASR large model stands out with its keen auditory perception and intelligent understanding capabilities, accurately recognizing speech even in complex environments, capable of recognizing 300 words per minute, which is faster than

On July 22, 01.AI held a product launch event in Beijing, officially launching the upgraded WanZhi Enterprise Large Model Platform 2.0 version, and simultaneously introducing an enterprise-level Agent intelligent entity, marking a significant step forward for AI technology in the enterprise application field. The newly released WanZhi Platform 2.0 and enterprise-level Agent aim to provide customized, high-performance AI solutions for enterprises through deep thinking and task planning capabilities. At the event, Dr. Li Kaifu, CEO of 01.AI, pointed out that there is a significant gap between foundational models and industrial applications at present.

On July 23, Alibaba Cloud officially announced that its latest AI programming large model, Qwen3-Coder, is fully open-sourced, which quickly sparked a buzz in the field of intelligent programming. Qwen3-Coder has achieved top performance among open-source models in agentic coding, agentic browser use, and basic coding tasks, thanks to its outstanding code generation and agent capabilities, marking a new stage in intelligent programming technology. The Qwen3-Coder model series provides

Today, iFLYTEK officially launched the intelligent office device X5. This world-first office device equipped with a local large model is now available on the market starting at 4999 yuan. The X5 features a 10.65-inch E Ink e-ink screen with a 300 PPI display resolution. The device weighs only 355 grams and has a thickness of 4.6 millimeters, becoming the lightest and thinnest e-ink device of its size in the world. The product comes in standard and Pro versions: the standard version includes a light gray leather cover and a gray body (6G+64G, 4999 yuan), while the Pro version offers a ink blue leather cover and a golden body,

NTU's 'WORLDMEM' model enhances virtual reality with long-term scene consistency via innovative memory mechanisms and conditional diffusion transformers.....

Recently, ByteDance's Seed team officially launched a new Vision-Language-Action Model (VLA) called GR-3. This model demonstrates breakthrough capabilities in the field of robotic manipulation, not only understanding language instructions that include abstract concepts, but also precisely handling flexible objects. It also has the ability to generalize quickly to new tasks and recognize new objects. This achievement is seen as an important advancement toward a 'general-purpose robot brain'. Traditional robotic manipulation models often rely on large amounts of robotic trajectory data for training.

01.AI launched 'Wanzi' enterprise Agent with 5 core functions: efficient execution, reliable results, self-learning, system integration, and rapid deployment. CEO Kai-Fu Lee highlights its co-creation model for AI-driven business transformation.....

ByteDance open sources the lightweight multilingual translation model Seed-X, which supports bidirectional translation in 28 languages and demonstrates performance comparable to top-level large models. This 7-billion-parameter model is based on the Mistral architecture and focuses on translation optimization, showing excellent performance in multiple areas. It uses an innovative training strategy to generate high-quality data and optimize deployment efficiency. This is another open-source project from ByteDance following BAGEL and Seed-Coder, promoting advancements in AI translation technology.

Alibaba's Qwen3-235B-A22B-Instruct-2507-FP8 update features non-thinking mode and 256K context length. Open-sourced on ModelScope/HuggingFace, enhancing complex AI tasks and text generation.....

OpenAI announced that GPT-5 will be launched soon, but clarified that it is not the experimental model that won the International Mathematical Olympiad. The CEO stated that the winning model used new technology and has mathematical capabilities far beyond current levels, while GPT-5, though expected to provide a surprising experience, will have different mathematical abilities. The community is热议 a suspected GPT-5 test model appearing on GitHub. OpenAI emphasized the need to distinguish between the actual capabilities of different models, providing clear guidance for market expectations. The AI field continues to closely watch the release of GPT-5 and technological breakthroughs. (140 characters)

FotographerAI introduces the revolutionary image generation framework ZenCtrl, which enables high-fidelity image generation with multiple perspectives and scenes from a single input image using GenAI technology. This framework breaks through three major pain points of traditional image generation: no need for fine-tuning, maintaining subject consistency, and supporting fine-grained control. Its modular toolkit integrates the entire workflow from preprocessing to postprocessing, built upon and optimized from OmniControl, with particular emphasis on enhancing subject retention capabilities. It supports diverse scenarios such as advertising creativity and virtual try-on, and future developments are also planned.