Search AI Products and News

The WHEE team under Meitu announced the official launch of its highly anticipated 'Ultra HD Video' feature. This feature is designed to help users repair and enhance video quality, whether it's long-forgotten old videos or blurry footage caused by shaky hands. Since the previous trial of the AI Ultra HD feature, WHEE has received a large amount of positive feedback and inquiries from users. Today, WHEE finally lived up to expectations by launching an upgraded 'Ultra HD Video' feature. Users just need to upload their videos and can activate the feature with one click.

The StepZen team announced that their latest generation foundational large model, Step3, is officially open-sourced. Step3 is a model designed for enterprises and developers who pursue the optimal balance between performance and cost, aiming to create the most suitable model for the inference era. The open-source address of this model includes GitHub, Hugging Face, and ModelScope, allowing developers to freely download and experience it. Step3 adopts an MoE architecture, with a total parameter count of 321 billion and an activated parameter count of 38 billion. It

The Qwen3-Coder series has welcomed a new member - Qwen3-Coder-Flash, a programming model affectionately called "a dessert-level" by developers. With its outstanding performance and efficient operation speed, this model brings new surprises to the field of programming. The full name of Qwen3-Coder-Flash is Qwen3-Coder-30B-A3B-Instruct. It performs excellently in terms of performance and efficiency, especially in agentic coding and browsing

Volc Engine announced that the daily tokens usage of Doubao Large Model exceeded 1.6 trillion, growing 137 times from last year, with a market share of 46.4%. At the launch event, new products such as Image Editing 3.0 and Simultaneous Interpretation 2.0 were introduced, and the large model was upgraded to version 1.6. President Tan Dai pointed out that AI is shifting from a tool to an intelligent agent, which will significantly change ways of living and working. IDC predicts that the public cloud large model usage in China will reach 114.2 trillion tokens in 2024. Volc Engine expects revenue to exceed 12 billion yuan in 2024. New products...

China's AI applications grow rapidly, with over 3.1B personal user registrations. In 2023, 3,755 global AI models emerged, 40% from China. By March, 346 AI services completed registration, with 80.9% usage in June. Industry scale is projected to exceed ¥700B in 2024, growing over 20% annually.....

The Wanxi Tianmu 2.0 large model by Wanxi Technology ranked fourth in China on the SuperCLUE global list, and was jointly recognized with Huawei's PanGu T2V, demonstrating its technological strength in the field of video generation. The company also established an AI video large model laboratory with Huawei Cloud, integrating resources to explore digital creative applications and promote the intelligent development of the industry. This collaboration marks an innovative breakthrough of AI technology in the field of video content, which is expected to enhance user experience and lead the industry direction.

Moonvalley's Marey model introduces Sketch-to-Video, enabling 1080p movie clips from text & sketches with 40% cost reduction. HBO-adopted, copyright-compliant, with future pro features.....

Midjourney adds 'For You' feature, offering personalized AI-generated content recommendations based on user interaction history (likes, moodboards). Located top-left, it enables one-click access to style-matched images/videos, adjustable via stylize/chaos parameters for precise genre alignment.....

Tencent's X-Omni AI model integrates reinforcement learning to unify image generation and understanding, outperforming rivals in text rendering and visual tasks without classifier guidance.....

The paper by Dr. Liang Wenfeng's team at DeepSeek in collaboration with Peking University has won the ACL 2025 Best Paper Award. This research proposes the Native Sparse Attention (NSA) mechanism, which achieves algorithm-hardware co-optimization through three parallel attention branches: compression, selection, and sliding. It improves the processing speed of long texts by 11 times and outperforms traditional full-attention models. NSA supports a context length of up to 1 million tokens, achieving an 11.6 times decoding speed improvement on 64k text, and a 27B parameter model demonstrates excellent performance on multiple benchmark tests.

The Qwen3-30B-A3B model has undergone a major update, launching the new version Qwen3-30B-A3B-Thinking-2507. This new version has achieved significant improvements in reasoning capabilities, general abilities, and context length, marking that the model is not only lighter but also has stronger thinking capabilities. Compared to the Qwen3-30-A3B model open-sourced on April 29, the new model has made progress in multiple key areas. In terms of mathematical abilities, Qwen3-30B-A3B-Thin

AMD Ryzen AI Max+395 upgrades to support 128B-parameter local AI models, requiring 128GB RAM (96GB VRAM). Runs 109B-parameter models like Meta Llama4Sout at 15 tokens/sec with 256K context length. MoE architecture enables efficient processing. 128GB AI workstations now priced at ~$1,800.....

Kunlun open-sources Skywork UniPic, a 1.5B-parameter multimodal model integrating image understanding, generation, and editing. It unifies multimodal tasks with end-to-end pretraining, offering strong performance comparable to larger models.....

Volcano Ark launches SeedEdit3.0, an AI image editor featuring voice-controlled editing. Based on Seedream3.0, it handles 10MP images with precise detail adjustments while preserving backgrounds. Excels in natural language understanding for complex edits like text replacement and day-night conversion. Applications include design, e-commerce, watermark removal, photo restoration, and style transfer, lowering editing barriers.....

Kunlun Wildfire releases the open-source multimodal pre-training model Skywork UniPic, integrating three functions: image understanding, text-to-image generation, and image editing. The model uses an autoregressive training approach, trained on large-scale high-quality data, and has strong generalization and transfer capabilities. Its open-source nature facilitates secondary innovation by developers and can be widely applied in content creation and industry scenarios. This is a significant move by Kunlun Wildfire in the AI field, which will promote the development and popularization of multimodal technology.

Li Auto launches i8, a 6-seat electric SUV with Pro/Max/Ultra variants (¥321,800-369,800), delivering Aug 20. Features: dual-motor AWD, 5C battery (670-720km range), LiDAR, VLA AI system, 9 airbags. Offers 3 interior colors, 20-inch wheels, ¥10k discount.....

Alibaba's Qwen team released Qwen3-30B-A3B-Instruct-2507, a 3B-parameter model rivaling Gemini2.5-Flash and GPT-4o. It excels in multilingual tasks, long-context processing, and benchmarks, with some metrics surpassing GPT-4o. Now available on ModelScope and HuggingFace.....

On July 30, Kunlun Infinite officially launched and open-sourced the autoregressive multi-modal unified pre-training model Skywork UniPic. The model integrates three core capabilities - image understanding, text-to-image generation, and image editing - in a single architecture. It demonstrates excellent generality and transferability through end-to-end pre-training on large-scale high-quality data.

CAS-SIAT's HYPIR model restores 8K images in 1.7s with advanced text reconstruction, outperforming traditional methods through optimized algorithms. Applied in photo restoration and heritage conservation.....

Google upgrades the Imagen4 image generation model. Imagen4Ultra ranks third globally in authoritative rankings, with performance close to GPT-4o and Seedream3.0. The new version significantly improves image details, realism, and style consistency, and handles complex prompts more accurately. It has a clear price advantage, with the standard version at $40 per thousand images and the Ultra version at $60 per thousand images, far lower than GPT-4o's $167. The generation speed is 9.5 seconds per image, faster than GPT-4o but slightly slower than Seedream3.