Versatile-OCR-Program is an open-source multimodal OCR tool tailored for educational scenarios and machine learning training. Combining technologies such as DocLayout-YOLO, Google Vision, and MathPix, it accurately recognizes text, mathematical formulas, and more...
Kimi-VL is a lightweight, multimodal visual language model open-sourced by Dark Side of the Moon. It is based on the lightweight MoE model Moonlight (16B total parameters, 2.8B activation parameters) and the native resolution MoonViT visual encoder (400M parameters)...
FantasyTalking is a novel framework jointly proposed by Alibaba's AMAP team and Beijing University of Posts and Telecommunications for generating realistic, animable virtual avatars from single static portraits. Based on a pre-trained video diffusion transformer model,...
BrowseComp is an open-source benchmark from OpenAI for evaluating the web browsing capabilities of AI agents. It contains 1266 highly challenging questions covering topics such as film, science and technology, art, history, sports, music, and more.
SenseNova V6 is SenseTime's sixth-generation multimodal fusion model series, based on a 600 billion parameter multimodal MoE architecture, enabling native fusion of text, images, and videos. SenseNova V6...
Mingqi is the first large-scale medical multimodal model for the precise diagnosis of rare diseases launched by the LoCCS Lab of the School of Computer Science at Shanghai Jiao Tong University. Based on a dual-engine architecture of "large model capability matrix + expert routing collaboration," it integrates medical imaging...
Fourier N1 is the first open-source humanoid robot released by General Robotics' Fourier Robotics division. Standing 1.3 meters tall and weighing 38 kilograms, it boasts 23 degrees of freedom and utilizes a composite structure of aluminum alloy and engineering plastics, combining strength and flexibility. ...
ScholarCopilot is an AI tool designed specifically for academic writing, developed by a research team at the University of Waterloo and Carnegie Mellon University in Canada. Based on the Qwen-2.5-7B model, it dynamically retrieves citations and performs joint optimization...
AgentPrune is a communication optimization framework for multi-agent systems driven by large language models (LLMs), jointly proposed by Tongji University, the Chinese University of Hong Kong, and other institutions. It uses a "pruning" technique to remove redundant or harmful communication content...
The HeartShadow Big Model is an AI big model developed by the HeartShadow Follows AI team, and has officially passed the national generative artificial intelligence service filing. The model focuses on the gaming field, deeply integrating game strategies and character companionship scenarios, and is comprehensively applied to "funny..."
FlexIP is a flexible subject attribute editing framework for image compositing proposed by Tencent, balancing identity preservation and personalized editing in image generation. The framework employs a dual-adaptor architecture to decouple identity preservation from personalized editing, through...
UNO is an innovative AI image generation framework launched by ByteDance, breaking through the limitations of traditional models in multi-subject generation. Through a "few-to-many" generalization method, it can generate high-quality single-subject and multi-subject images, solving the problems in multi-subject scenes...
Skywork-OR1 (Open Reasoner 1) is an open-source, high-performance inference model series launched by Kunlun Tech. These models overcome the limitations of large models in logical reasoning and solving complex tasks. Skywork-OR1 includes three models...
Seed-Thinking-v1.5 is an inference intelligence model launched by ByteDance. It adopts a hybrid expert (MoE) architecture with a total of 200 parameters, activating 20 parameters per iteration. The model performs excellently in multiple benchmark tests, such as AI...
HoloPart is a novel diffusion model developed by the University of Hong Kong and the VAST team. It supports the decomposition of 3D objects into complete, editable semantic parts, even when the parts are occluded.
UniRig is an innovative automatic skeleton rigging framework jointly developed by the Department of Computer Science at Tsinghua University and VAST, used to handle complex and diverse 3D models. Based on a large autoregressive model and a bone point cross-attention mechanism, it generates high-quality...
The AI Scientist-v2 is a fully automated end-to-end artificial intelligence system developed by Sakana AI, the University of British Columbia, the Vector Institute, and other institutions. It supports the autonomous generation of scientific hypotheses...
PixelFlow is an image generation model jointly developed by the University of Hong Kong and Adobe, supporting the generation of images directly in pixel space. PixelFlow is based on efficient cascaded flow modeling, progressively improving from low to high resolution, significantly...
GPT-4.1 is OpenAI's latest next-generation language model, comprising three versions: GPT-4.1, GPT-4.1 mini, and GPT-4.1 nano. This series of models shows significant improvements in encoding capabilities, instruction compliance, and long text processing...
The Agent Development Kit (ADK) is Google's first open-source AI agent development toolkit. It's a code-first Python toolkit that simplifies the construction, evaluation, and deployment of complex AI agents. ADK supports multiple agents...
MiniMax MCP Server is a multimodal generation server based on the Model Context Protocol (MCP) developed by MiniMax Technology. Through simple text input, it can invoke video generation, image generation, speech generation, and audio generation...
BiliNote is an open-source AI video note-taking assistant that supports automatically extracting video content from video links such as Bilibili and YouTube to generate well-structured and concise Markdown format notes.
Keling 2.0 is Kuaishou's next-generation AI video generation model, now available on the Keling AI Video Platform. The platform update includes two basic models: Keling 2.0 (Master Edition) and Ketu 2.0. Ketu 2.0 significantly improves image generation quality...
MedReason is a medical reasoning framework developed by institutions such as the University of California, Santa Cruz, the University of British Columbia, and Nanyang Technological University in Singapore. It leverages knowledge graphs to enhance the application of large language models (LLMs) in medical...