Search AI Products and News

Poe launches the API feature, allowing developers to directly call all AI models and robots on the platform through subscriptions, including image and video processing capabilities. The API is compatible with the OpenAI interface, requiring no additional learning costs, and supports diverse needs such as text generation and creative content creation. The subscription model lowers the entry barrier, enabling developers to quickly integrate Poe's technological capabilities and explore the potential of AI applications. This update marks Poe's transformation from a chat platform to a developer tool.

xAI's Grok AI introduces 'Imagine' for 6s videos with sound, including explicit content, raising deepfake concerns. Part of Grok4 update, already controversial. 84% Americans support banning non-consensual deepfake porn. Critics warn of sexual exploitation risks, but launch proceeds for October.....

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.....

Meta upgrades the Avatar system, adding more than 10 preset body types and detailed area adjustment features, supporting 18 postures and over 50表情 stickers. The AI styling assistant can quickly generate virtual avatars based on text descriptions, currently available only in the US and Canada. Parametric facial control enables billions of combination possibilities, allowing users to precisely customize their virtual avatars.

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

LiblibAI's Lovart AI launches globally as the first commercial AI design Agent, serving 800K users across 70 countries. It integrates top models via a 'Creative Reasoning Engine' to generate 40 professional designs from text, covering branding and marketing. Features include ChatCanvas for natural language adjustments and learning user preferences. A China-optimized 'Stellar Agent' enhances Chinese semantic processing. The tool lowers design barr....

ByteDance launched the Trae IDE 2.0 version, which realizes a breakthrough in AI autonomous programming with its revolutionary SOLO mode. This mode integrates tools such as the editor and terminal, enabling the completion of the entire development process from concept to deployment without continuous human intervention. Tests show that SOLO can autonomously develop and deploy applications, such as a WeChat Moments analysis tool. Compared to traditional AI coding tools, Trae 2.0 provides a more complete end-to-end solution, surpassing similar competitors. Currently, Pro users can experience this feature, which requires an invitation code to unlock. This upgrade marks a significant step forward.

Zhuji Power has received strategic investment from JD.com, accelerating the mass production of humanoid robots and the research and development of embodied intelligence technology. The company has launched the IDS Collaborative Innovation Strategy, serving innovators, developers, and system integrators by providing an open technology platform. A full-size humanoid robot is expected to be launched in 2025, with the ability to adapt to complex environments. The bipedal robot TRON1 has already been applied in areas such as inspection and security, and a developer kit has been released. The company's LimX VGM large model enables cross-platform operation capabilities, and more visual-language-action large model results will be released this year.

Lightricks releases open-source LTX-Video13B, a 13B-parameter video generation model with multi-scale rendering, achieving 30x faster speeds. It runs on consumer GPUs, supports 1216×704 real-time generation, and offers text/image/video-to-video modes. The model enhances coherence and detail, enabling keyframe control and style transfer. Free for SMEs, it includes training tools and optimized versions to democratize AI video creation.....

Three US authors sue AI firm Anthropic for using copyrighted books to train models. Partial dismissal granted, but piracy claims proceed. Case highlights AI-copyright tensions amid rising lawsuits from creators like Reddit, though some opt for profit-sharing deals.....

Anthropic sued by authors for allegedly using pirated books to train AI chatbot Claude. Case proceeds as class action, highlighting copyright tensions in AI industry.....

MiniMax launches a revolutionary AI development tool called MiniMax Agent, allowing users to automatically generate a complete e-commerce website with just natural language descriptions, including front-end, back-end, database, and payment interface functionalities. This technology significantly lowers the development barrier, supports international payments, real-time data updates, and automatic deployment, making it especially suitable for start-ups and individual entrepreneurs. This innovation will change traditional development models, making it possible to rapidly bring ideas to life and offering new opportunities in the e-commerce field.

BAAI released RoboBrain2.0 (32B) and RoboOS2.0 framework. RoboBrain2.0 excels in spatiotemporal cognition and complex tasks, while RoboOS2.0 is the first embodied AI SaaS framework supporting lightweight deployment and multi-robot collaboration. Both are now open-source.....

Meta and Berkeley developed StreamDiT for real-time AI video generation: 1) 16fps 512p on single GPU, 4B-param model creates 1-min videos with live edits; 2) Novel buffer enables parallel processing (2 frames/0.5s) with 8-step optimization; 3) Trained on 3K HD videos + 2.6M dataset; 4) Outperforms rivals in motion smoothness, 30B-param version shows quality potential; 5) Enables interactive video despite transition flaws.....

Alibaba open sources the audio generation model ThinkSound, which supports multimodal inputs such as video, text, and audio, and can automatically generate high-fidelity sound effects that highly match the visuals. The model uses chain reasoning technology to achieve precise synchronization between audio and video, and is applicable to fields such as film and games. As an open source project, ThinkSound lowers the barriers to sound effect creation, and developers can freely access it through multiple platforms. This is Alibaba's latest breakthrough in the field of multimodal AI, and will drive the development of sound generation technology.