Hunyuan Image-Generated Video is an open-source image-generated video model launched by Tencent Hunyuan. Users can upload an image and add a brief description to generate a 5-second short video that animates the image. The model supports lip-syncing, motion-driven editing, and background sound effects...
Granite 3.2 is IBM's open-source multimodal AI model family, possessing powerful inference, visual understanding, and prediction capabilities. Granite 3.2 includes multiple versions, and the Granite 3.2 Instruct provides experimental chained inference capabilities...
Resume Matcher is an open-source AI resume optimization tool that helps job seekers increase the chances of their resumes passing automated screening systems (ATS). Resume Matcher analyzes resumes and job descriptions to extract key skills, experience, etc.
OpenManus is an open-source fork of Manus released by the MetaGPT team, providing an AI agent that requires no invitation code. OpenManus is based on a modular design, supports multiple language models and toolchains, and can execute code, process files, etc.
NEXUS-O is a multimodal AI model developed by HiThink Research Institute, Imperial College London, Zhejiang University, Fudan University, Microsoft, Meta AI, and other institutions. It enables comprehensive perception and interaction of language, audio, and visual information...
OWL (Optimized Workforce Learning) is a multi-agent collaborative system developed based on the CAMEL-AI framework. It achieves efficient task automation through dynamic interactions between agents. This is accomplished through role assignment and task decomposition...
TicVoice 7.0 is Mobvoi's seventh-generation high-quality TTS (text-to-speech) engine, based on the next-generation speech generation model Spark-TTS. TicVoice 7.0 utilizes the innovative BiCodec encoding method to segment speech...
GEN3C is a new generative video model developed by NVIDIA, the University of Toronto, and the Vector Institute. It generates high-quality 3D video content based on precise camera control and spatiotemporal consistency. GEN3C builds a point cloud-based 3D cache...
Mercury Coder is Inception Labs' first diffusion-based large language model (dLLM), and is a model in the Mercury series specifically designed for code generation. Mercury Coder is based on a coarse-to-fine generation approach...
GCDance (Genre-Controlled 3D Full Body Dance Generation Driven by Music) is a 3D dance generation framework developed by the University of Surrey and Jiangnan University in the UK. It can generate dances that conform to music and text prompts...
LaWGPT is a large-scale Chinese legal language model developed by Nanjing University. It is based on the LLaMA model and undergoes secondary pre-training, incorporating a large amount of Chinese legal knowledge. Focusing on the legal field, it can understand and generate legal-related texts and is applicable to...
autoMate is a local automation tool based on AI and RPA that automates complex tasks using natural language. Based on a large language model, autoMate allows users to describe tasks in natural language without needing programming knowledge, enabling them to complete tasks...
URO-Bench is a comprehensive benchmarking tool for end-to-end speech dialogue models (SDMs). It covers multi-dimensional tasks such as multilingualism, multi-turn dialogue, and paralinguistic information to comprehensively evaluate the performance of speech dialogue models.
Nanobrowser is an open-source Chrome extension focused on AI-driven web page automation. Nanobrowser utilizes a multi-agent system to perform complex web page tasks, such as information extraction and automated operations. Users can...
DINO-XSeek is a multimodal object detection model developed by IDEA Research, combining visual perception and natural language understanding capabilities. DINO-XSeek accurately locates objects in images based on complex linguistic descriptions and identifies object attributes...
AppAgentX is a new self-evolving GUI (Graphical User Interface) agent framework developed by Westlake University. It improves the efficiency and intelligence of agents in smartphone interactions by abstracting high-level actions from execution history. AppAgent...
GO-1 (Genie Operator-1, the first universal embodied base model launched by Zhiyuan Robotics. The model adopts the Vision-Language-Latent-Action (ViLLA) architecture, consisting of VLM (Vision-Language-Latent-Action)...
Gemini Embedding is an advanced text embedding model from Google that transforms text into high-dimensional numerical vectors to capture its semantic and contextual information. Gemini Embedding is trained on the Gemini model and has...
DoraCycle is a unified generative model for multimodal domain adaptation developed by the Show Lab at the National University of Singapore. It achieves different... through two multimodal loops (text-to-image-to-text and image-to-text-to-image).
IMAGPose is a unified conditional framework for human pose-guided image generation, developed by Nanjing University of Science and Technology. It addresses the limitations of traditional methods in pose-guided human image generation, such as the inability to simultaneously generate multiple different poses...
Heygem is an open-source digital human model developed by Silicon Intelligence, designed specifically for Windows systems. Based on advanced AI technology, it can clone a digital human's image and voice within 30 seconds using only a 1-second video or a single photograph.
LanDiff is an innovative hybrid framework for high-quality text-to-video (T2V) generation. It combines the advantages of autoregressive language models (LLM) and diffusion models, employing a coarse-to-fine generation approach...
FoxBrain is a large-scale Traditional Chinese language model (LLM) developed by Foxconn Research Institute, boasting powerful reasoning capabilities. Based on the Meta Llama 3.1 architecture, it has 70 parameters and focuses on mathematical and logical reasoning. FoxB...
COMET, developed by ByteDance, is an optimization system for Mixture-of-Experts (MoE) models, designed to address the issue of excessive communication overhead in distributed training. Based on fine-grained computation-communication overlap technology, COMET...