Shadow is an open-source AI programming agent that helps developers understand, reason, and contribute to existing codebases. Shadow supports GitHub repository integration, can generate pull requests, manage branches, and provide real-time task status updates. ...
MoE-TTS is the first role-description speech synthesis framework based on MOE launched by the Kunlun Wanwei speech team, specifically designed to improve the understanding of open-domain text descriptions. The model utilizes a Mixture-of-Experts (MOE) approach...
DINOv3 is a general-purpose, state-of-the-art (SOTA) visual foundation model from Meta. Trained on unlabeled data, it generates high-quality, high-resolution visual features, suitable for multiple tasks such as image classification, semantic segmentation, and object detection.
Genie Envisioner is the first open-source robot world modeling platform launched by Logic Studio. The platform integrates policy learning, evaluation, and simulation functions through a unified video generation framework. Core components include GE-Base (a large-scale...
VeOmni is an open-source, full-modal distributed training framework developed by ByteDance's Seed team, based on PyTorch. VeOmni is model-centric, decoupling distributed parallel logic from model computation and supporting flexible combinations of various parallel...
OpenDeRisk is an AI-native risk intelligence management system that provides comprehensive 24/7 protection for application systems. Based on Deep Research RCA, the system quickly locates the root cause of problems, combined with...
MemU is an open-source AI memory framework for AI-powered emotional companionship. MemU can remember every conversation between a user and AI, extract key points, and build a knowledge graph, enabling AI to truly understand the user.
Chaterm is an open-source, intelligent, and secure AI endpoint tool designed for cloud professionals and DevOps engineers. Based on an AI Agent, it allows users to use natural language instead of complex command-line operations. Chaterm provides access based on personal knowledge...
AudioGenie is a multimodal audio generation tool developed by Tencent AI Lab. It can generate various audio outputs, including sound effects, speech, and music, from multiple modal inputs such as video, text, and images. The tool employs a training-free multi-agent approach...
Qwen-Image-Edit is a versatile image editing model built on the Qwen-Image architecture with 20 billion parameters. The model possesses both semantic and visual editing capabilities, enabling low-level visual editing (such as adding, deleting...).
Open-Lovable is an open-source project from the Firecrawl team that uses AI technology to quickly clone any website into a modern React application. After the user enters the URL of the target website, Firecrawl crawls the content...
Klear-Reasoner is a reasoning model launched by Kuaishou based on Qwen3-8B-Base, focusing on improving mathematical and code reasoning capabilities. The model is trained through long thought chain supervised fine-tuning (long CoT SFT) and reinforcement learning (RL)...
Nano Banana is an AI image generation and editing model launched by Google. The Gemini 2.5 Flash Image (codenamed Nano Banana) model has extremely high facial consistency, accurately reproducing facial features and expressions.
CombatVLA is a high-efficiency visual-verbal-motion (VLA) model developed by the Future Life Lab team at Taotian Group, specifically designed for combat missions in 3D action role-playing games (ARPGs). The model is based on a 3B parameter scale and...
The NVIDIA Nemotron Nano 2 is a high-performance inference model from NVIDIA, with 9 bytes of parameters. The model is based on a hybrid Mamba-Transformer architecture, pre-trained on 20 trillion tokens, and supports a 128k context length.
DeepSeek V3.1 is the latest AI model version released by DeepSeek, featuring a hybrid inference architecture that allows for seamless switching between thinking and non-thinking modes, significantly improving thinking efficiency. The model has undergone several upgrades compared to V3, including...
AutoCodeBench is a benchmark dataset launched by Tencent Hunyuan specifically for evaluating the coding capabilities of large-scale programming models. It contains 3920 questions, evenly distributed across 20 programming languages. The dataset is characterized by its high difficulty, practicality, and diversity, and can...
Seed-OSS is a series of large-scale language models open-sourced by ByteDance's Seed team, focusing on long text processing, reasoning, and intelligent agent capabilities. The model includes multiple versions, such as Seed-OSS-36B-Base and Seed-OSS-36B-I...
ToonComposer is a generative AI tool jointly developed by researchers from the Chinese University of Hong Kong, Tencent PCG ARC Lab, and Peking University. It can transform sketches into professional-grade animations in seconds. ToonComposer is based on generative post-processing...
Intern-S1-mini is a lightweight, open-source multimodal inference model developed by the Shanghai Artificial Intelligence Laboratory. It is built using the same technology as Intern-S1. The model integrates an 8-bit dense language model (Qwen3) and a 0.3-bit visual encoding...
Fun-ASR is a next-generation speech recognition model jointly developed by DingTalk and the Tongyi Labs speech team. Trained on massive amounts of audio data, it can accurately recognize professional terms from more than ten industries, including internet, technology, home decoration, and animal husbandry...
Grok 2.5 is an open-source artificial intelligence model from xAI, a company founded by Elon Musk. The model files, available for download on Hugging Face, consist of 42 files, totaling approximately 500GB in size, and require eight GPUs with over 40GB of dedicated memory each to run...
Prompt Optimizer is an open-source AI prompt optimization tool that helps users quickly write high-quality prompts to improve the quality of AI output. It offers one-click prompt optimization, supports multiple rounds of iterative improvement, and enhances the accuracy of AI responses...
ComoRAG is a cognitive heuristic retrieval augmented generation (RAG) framework jointly launched by the Institute of Future Technology at South China University of Technology, the WeChat AI team, and other institutions. It is specifically designed for understanding and reasoning about long narrative texts.