Search AI Products and News

Qafind Labs recently released its newly developed ChatDLM model, an innovative achievement that has drawn widespread attention in the field of artificial intelligence. ChatDLM is the first model to deeply integrate "Block Diffusion" and "Mixture of Experts (MoE)". It achieves an astonishing inference speed of 2,800 tokens/s on GPU, supporting an ultra-large context window of 131,072 tokens, ushering in a new era of document-level generation and real-time conversation.

On April 27, Zhipu AI (Z.ai) and Shengshu Technology (shengshu.com), two leading artificial intelligence companies under Tsinghua University, announced a major strategic partnership. This collaboration aims to leverage both companies' technological expertise in large language models and multi-modal generative models to jointly advance the technological innovation and industrial application of domestic large models.

In today's increasingly competitive AI landscape, Ema has launched a new language model, EmaFusion, claiming to surpass several well-known AI models, including O3, Gemini, and Sonnet, in both cost and accuracy. Unlike traditional single-strategy systems, EmaFusion employs a "cascaded" judgment system that dynamically balances cost and accuracy, while also allowing users to fine-tune it based on specific task requirements. Ema's CEO, Sur...

Moonshot AI recently announced the launch of Kimi-Audio, a new open-source audio foundation model aimed at advancing the field of audio understanding, generation, and interaction. This release has garnered significant attention from the global AI community and is considered a major milestone in the development of multimodal AI. This report provides a comprehensive overview of Kimi-Audio's core features, performance, and industry impact. Breakthrough Features: Versatile Audio Processing Capabilities Kimi-Audio-7B-Instruct based on Qwen

Step1X-Edit is a groundbreaking open-source image editing model that achieves performance comparable to leading closed-source models such as GPT-4o. It offers a powerful and versatile solution for various image manipulation tasks.

Shanghai is actively promoting innovation and upgrading within its automotive industry, focusing on the application of advanced technologies including high-performance computing chips and intelligent driving large models. This initiative aims to enhance the competitiveness and technological advancement of the city's automotive sector.

Google DeepMind has officially released its latest music generation model, Lyria2, marking another significant breakthrough in artificial intelligence for music creation. This new model, with its high-fidelity audio generation and professional-grade sound quality, provides musicians, producers, and creators with more powerful creative tools. Lyria2: High-fidelity sound quality, capturing the subtle beauty of music. Lyria2 represents DeepMind's latest advancement in music generation technology, offering significant improvements in sound quality and creative flexibility compared to previous models.

3DV-TON (Textured 3D-Guided Consistent Video Try-on via Diffusion Models) is an innovative technology that delivers a realistic video try-on experience with texture 3D guidance using diffusion models. According to AIbase, 3DV-TON leverages advanced 3D geometry and texture modeling combined with video diffusion models to ensure consistency and realism of clothing in dynamic videos, offering significant potential for e-commerce, fashion, and virtual reality applications.

Ele.me has officially launched Xiaoe, China's first AI-powered intelligent agent for delivery riders, leveraging AI technology to restructure the rider's work environment and achieve improvements in both operational efficiency and safety. This feature has initially launched in four cities: Wuxi, Shenyang, Foshan, and Suzhou, covering delivery riders using the Fengniao Zhonbao app.

Recently, Feishu and Shengshu Technology reached a strategic partnership, officially integrating their domestic video large model, Vidu, into Feishu's multi-dimensional table function. Feishu users can now easily call the Vidu video large model through AI field shortcuts in multi-dimensional tables, improving work efficiency and enriching data processing capabilities. Vidu, from Shengshu Technology, is a powerful video large model capable of efficient video generation and processing. This collaboration aims to enable Feishu users to more conveniently utilize AI technology for video-related tasks.

Bytedance's Volcano Engine announced the full launch of its newly released Doubao 1.5 Deep Thinking model on the edge large model gateway, offering users up to 5 million free tokens. This move has garnered significant attention in the AI field.

At the Baidu Create Developer Conference on April 25th, Baidu founder Robin Li unveiled the latest additions to the Ernie family of large language models: Ernie 4.5 Turbo and X1 Turbo. These new models boast significant improvements in speed and cost-effectiveness, marking a major advancement in Baidu's AI capabilities. Ernie 4.5 Turbo, in particular, offers a dramatic speed increase and higher processing efficiency while boasting an impressive 80% price reduction compared to its predecessor. Specifically, the cost per million tokens...

In the field of artificial intelligence, Meta recently introduced the WebSSL family of models. These models, ranging in size from 300 million to 7 billion parameters, are trained on purely image data and aim to explore the vast potential of language-free visual self-supervised learning (SSL). This new research opens up new possibilities for future multimodal tasks and offers a fresh perspective on understanding how visual representations are learned. Previously, OpenAI's CLIP model was known for its performance in multimodal tasks such as visual question answering (VQA) and document understanding.

A GitHub open-source project named "system-prompts-and-models-of-ai-tools" has gained significant traction, accumulating 30.5K stars and becoming a popular resource for AI developers and researchers. This project, according to AIbase, compiles system prompts and model configurations for 9 mainstream AI tools, encompassing over 6500 lines of content and covering v0, Cursor, Manus, Same.dev, Lovable, Devin, and Rep.