Async is an open-source developer tool that provides experienced developers with a unified workflow that combines AI coding, task management, and code review, helping them improve their productivity through AI technology.
InternVL3.5 (Shusheng·Wanxiang 3.5) is an open-source multimodal large model from the Shanghai Artificial Intelligence Laboratory. The model has undergone comprehensive upgrades in general-purpose capabilities, inference capabilities, and deployment efficiency, offering nine sizes ranging from 1 billion to 241 billion parameters...
MAI-1-preview is an end-to-end trained foundational model developed by Microsoft's AI team, providing users with the ability to follow instructions and answer everyday queries.
MobileCLIP2 is a high-performance edge multimodal model developed by Apple researchers, and is an upgraded version of MobileCLIP. It features optimizations in multimodal reinforcement training, achieving superior CLIP performance through training on the DFN dataset...
Step-Audio 2 mini is an open-source end-to-end speech model released by Step-Audio. Breaking away from traditional speech model structures, it adopts a true end-to-end multimodal architecture, directly converting raw audio input into speech response output with lower latency...
AudioStory is an audio generation technology released by Tencent ARC Labs, capable of generating high-quality, long-form narrative audio based on natural language descriptions. It employs a divide-and-conquer strategy, breaking down complex narrative requests into ordered subtasks, and then...
Open-Fiesta is an open-source, multi-model AI chat platform built on Next.js 14. The platform supports various AI providers (such as Gemini and OpenRouter) and multiple models, allowing users to select up to five models simultaneously...
AgentScope 1.0 is an open-source multi-agent development framework from Alibaba Tongyi. Through a three-layer technical architecture—the AgentScope core framework, AgentScope Runtime, and AgentScope Studio—it provides a complete development and deployment solution...
HunyuanWorld-Voyager (abbreviated as HunyuanVoyager) is Tencent's first ultra-long roaming world model supporting native 3D reconstruction. It's a novel video diffusion framework capable of generating user-defined camera paths from a single image...
USO (Unified Style-Subject Optimized) is a unified framework for decoupling and recombining content and style, developed by ByteDance's UXO team. It allows for the free combination of any theme and style in any scenario to generate content with high...
Prompt Tools is an open-source desktop application that helps users efficiently manage AI-generated prompts. The tool supports cross-platform operation (Windows, macOS, Linux) and provides a fast and secure experience.
The Coding Interview Problem Solver is a programming aid tool for Chinese users. It can capture programming problems on the screen using keyboard shortcuts and generate solutions and code in real time with the help of AI.
RealDevWorld is a next-generation automated testing tool developed by the MetaGPT team. Based on a multi-agent framework, it simulates the workflow of a real development team, integrating requirements analysis, test case generation, code debugging, and final...
Kimi K2-0905 is the latest AI model from Moonlit Dark Side Technology Co., Ltd. It boasts powerful programming assistance capabilities, excelling in code generation and front-end development. Its context length has been extended to 256K, supporting complex...
Apertus is Switzerland's first large-scale, open, multilingual language model, jointly developed by EPFL, ETH Zurich, and the Swiss National Supercomputing Center (CSCS). It has two parameter versions: 70B and 8B.
Super Agent Party is an open-source 3D AI desktop companion software that integrates multiple functions such as desktop pet, intelligent assistant, knowledge base, and robot control. Users can quickly deploy it to Windows, macOS, etc. via Docker or source code.
InfinityHuman is a commercial-grade long-sequence audio-driven human video generation model launched by a joint team from ByteDance and Zhejiang University, opening a new chapter in the practical application of AI digital humans.
mobile-use is a mobile AI Agents tool that supports natural language control of Android and iOS devices, helping users operate their phones. Users issue commands in everyday language, and the tool automatically completes tasks such as opening apps, filling out forms, etc.
EmbeddingGemma is an open-source multilingual text embedding model from Google, designed specifically for edge AI and supporting deployment on devices such as laptops and mobile phones. The model has 308 million parameters, is built on the Gemma 3 architecture, and supports over 100 languages...
OneCAT is a new unified multimodal model launched by Meituan. It employs a pure decoder architecture and seamlessly integrates multimodal understanding, text-to-image generation, and image editing functions. The model abandons the reliance on external vision in traditional multimodal models...
Qwen3-Max-Preview is the latest flagship large-scale language model released by Alibaba's Tongyi Qianwen team. It boasts the largest number of parameters in the Qwen3 series, exceeding one trillion. The model excels in inference, instruction following, ...
Claudable is an open-source web application builder based on Next.js. It combines the advanced AI agent capabilities of Claude Code and Cursor CLI with Lovable's simple and intuitive application building experience. Users only need to describe...
FineVision is an open-source visual language dataset released by Hugging Face, used to train advanced visual language models. It contains 17.3 million images, 24.3 million samples, 88.9 million dialogue rounds, and 9.5 billion answer tags. ...
Flowith is powered by top-tier large models and supports a variety of advanced AI models (such as Grok4, GPT-5, Claude, etc.) and multimodal image and video models (Veo 3, Midjourney, SDXL). Flowith supports text creation, image editing, and more.