RoboBrain is an open-source embodied brain model launched by the Beijing Academy of Artificial Intelligence (BAAI), propelling single-machine intelligence towards swarm intelligence. It consists of three modules: a base model for task planning, an A-LoRA module for operational area awareness, and...
RoboOS is the first cross-ontology embodied cerebellum-cerebellum collaborative framework launched by the Beijing Academy of Artificial Intelligence. Based on a hierarchical "brain-cerebellum" architecture, the embodied brain RoboBrain is responsible for global perception and decision-making, while the cerebellum skill library is responsible for low-latency and precise execution,...
Vidu Q1 is a highly controllable video model developed by the team of Professor Zhu Jun, Vice Dean of the Institute for Artificial Intelligence at Tsinghua University and founder and chief scientist of Shengshu Technology. It supports the generation of 1080p high-definition video with delicate image quality and rich details...
ObjectMover is a novel image editing model jointly proposed by the University of Hong Kong and Adobe Research. It addresses issues such as inconsistent lighting and shadows, and object distortion that arise when moving, inserting, and removing objects in an image. It...
PhysGen3D is an innovative framework that transforms single images into interactive 3D scenes, generating physically realistic videos. It combines image-based geometric and semantic understanding with physically based simulation, enabling the creation of interactive 3D scenes from single images...
EmotiVoice is an open-source text-to-speech (TTS) system from NetEase Youdao, offering multilingual, multi-voice, and prompt-controlled audio. EmotiVoice supports English and Chinese, provides over 2000 voices, and supports generating text with different emoticons based on prompts...
Vibe Draw is an open-source AI 3D modeling tool that transforms user-drawn sketches on a 2D canvas into beautiful 3D models. Users can use text prompts or continue drawing and iteratively refining the model, and export it to a standard format with one click (...
OmniSQL is an open-source text-to-SQL model that efficiently converts natural language problems into SQL queries. It uses an innovative data synthesis framework to generate the first million-level text-to-SQL dataset, SynSQL-2.5M, which includes...
MegaTTS 3 is a zero-shot text-to-speech synthesis system developed by ByteDance in collaboration with Zhejiang University. It employs a lightweight diffusion model with only 0.45 bytes of parameters, enabling efficient generation of high-quality speech. The system decomposes speech into content, audio, and...
AutoGLM is the first free AI agent launched by Zhipu, possessing in-depth research and operational capabilities. It can simulate human thought processes and handle complex open-ended questions. AutoGLM boasts powerful reasoning abilities and autonomous operation...
cpmGO (MiniCPM Super Assistant) is the world's first purely edge-based intelligent assistant launched by Mianwall Intelligence, designed specifically for automotive intelligent cockpits. cpmGO is developed based on the Mianwall MiniCPM edge model and features vision, voice, and multi-functionality...
AReaL-boba is an open-source reinforcement learning training framework jointly developed by Ant Research Institute and Tsinghua University. AReaL-boba is an upgraded version of AReaL, lowering the barrier to entry for reinforcement learning training, allowing users to easily train inference models.
Qlib is an AI-powered quantitative investment tool developed by Microsoft Research Asia for the financial industry. It helps quantitative researchers explore the potential of AI technology in the investment field. Qlib provides high-performance data processing infrastructure to support data extraction...
OpenDeepSearch is an open-source deep search tool that improves search performance based on open-source inference models and inference agents. It is specifically optimized for seamless integration with Hugging Face's SmolAgents and supports deep web search and...
Hi3DGen is a high-fidelity 3D geometry generation framework jointly developed by researchers from the Chinese University of Hong Kong, Shenzhen, ByteDance, and Tsinghua University. It can generate high-fidelity 3D models from 2D images, using normal maps as an intermediate representation...
GLM-Z1-Air is a deep thinking model developed by Zhipu Technology, based on GLM-4-Air-0414. GLM-Z1-Air incorporates more inference data during the pre-training phase and deeply optimizes its general capabilities during the alignment phase, demonstrating powerful mathematical...
AI-ClothingTryOn is a Python-based desktop application that uses Google Gemini AI technology to implement virtual try-on functionality. AI-ClothingTryOn supports uploading separate photos of the person and clothing, based on AI technology...
GLM-4-Air-0414 is a base model with 32 billion parameters launched by Zhipu Technology, and is the model behind AutoGLM. GLM-4-Air-0414 incorporates more code and inference data during the pre-training phase, targeting intelligent physical abilities...
Second Me is an open-source AI identity model developed by Mindverse. It supports the creation of fully private and deeply personalized AI agents that represent a user's "true self." Second Me offers Chat Mode and Brid...
FinGPT is an open-source, large-scale pre-trained language model for the financial field, driving financial innovation based on natural language processing technology. Through reinforcement learning and human feedback (RLHF) techniques, it can learn individual preferences and provide personalized investment recommendations...
GLM-Z1-Rumination is a contemplation model developed by Zhipu, further optimized from GLM-Z1, with specific parameters GLM-Z1-Rumination-32B-0414. GLM-Z1-Rumination is trained using extended reinforcement learning...
Runway Gen-4 is a high-fidelity AI video generation model developed by the AI startup Runway. It supports maintaining consistency between people, objects, and the environment across multiple scenes without requiring additional fine-tuning or training. Users provide a reference image...
Neo-1, launched by VantAI, is the world's first AI model to unify de novo molecule generation and atomic-level structure prediction. It can predict biomolecular structures, generate entirely new molecules, and has shown promise in designing novel therapeutics such as molecular gels...
Amazon Nova Act is a general-purpose AI agent from Amazon AGI Labs, used to perform tasks within a web browser. Amazon Nova Act allows developers to build prototype intelligent agent applications based on the accompanying SDK, enabling tasks such as...