Hunyuan3D 2.0 is a large-scale 3D asset generation system launched by Tencent, focusing on generating high-resolution 3D models from text and images. The system employs a two-stage generation process: first, it generates a textureless geometric model, then...
Claude 3.7 Max is Cursor's latest AI model designed specifically for complex coding tasks, touted as the most powerful code assistant ever. Based on the Claude 3.7 thinking model, it features a massive 200k context window and supports processing...
Skywork R1V is Kunlun Tech's first open-source industrial multimodal thought chain reasoning model, possessing powerful visual chain reasoning capabilities. Skywork R1V can perform multi-step logical reasoning on visual input, solving complex visual tasks...
Chirp 3 is a high-definition speech synthesis model from Google Cloud, designed to generate natural and vivid speech. It supports 248 voices and 31 languages, capturing subtle differences in human intonation for more realistic speech output...
LangManus is an AI automation framework based on a hierarchical multi-agent system design. It includes various agents, such as coordinators, planners, researchers, and programmers, each with its own role, collaborating to complete complex tasks. The framework supports multiple...
Cube 3D is an AI-powered 3D generator from Roblox, efficiently generating 3D models and environments using AI technology. Cube 3D supports generating complete 3D objects directly from text descriptions and is compatible with game engines. Cube 3D can...
SmolDocling (SmolDocling-256M-preview) is a high-efficiency, lightweight, multimodal document processing model. It can convert document images end-to-end into structured text, supporting the recognition of various elements such as text, formulas, and charts...
The GR00T N1 is NVIDIA's first open-source foundational model, designed specifically for general-purpose humanoid robots. It enables maneuvering tasks in diverse environments based on multimodal inputs (such as language and images). The GR00T N1 is based on large-scale...
UniFluid is a unified autoregressive framework jointly developed by Google DeepMind and MIT for joint visual generation and understanding tasks. It processes multimodal image and text inputs based on continuous visual tags to generate discrete...
Multi-Speaker is the world's first high-resolution multi-speaker separation model launched by AudioShake. It supports the precise separation of multiple speakers in an audio stream into different tracks, solving the problem of traditional audio tools handling overlapping speech...
UniAct is a novel embodied fundamental model framework that addresses the problem of behavioral heterogeneity among different robots. By learning common behaviors, it captures shared atomic behavioral features across different robots, eliminating the impact of differences in physical form and control interface...
ReCamMaster is a video re-rendering framework jointly developed by Zhejiang University, Kuaishou Technology, and others. It can regenerate video content based on new camera trajectories. Through pre-trained models and frame-dimensional conditional mechanisms, combined with multi-camera synchronization data...
Stable Virtual Camera is an AI model from Stability AI that can convert 2D images into 3D video with realistic depth and perspective. Users can specify camera paths and various dynamic paths (such as...)
Instella is an open-source language model series with 3 billion parameters, developed by AMD. The model was trained entirely from scratch on an AMD Instinct™ MI300X GPU, based on an autoregressive Transformer architecture, and includes 36 decoder layers and 3...
Maestro is an end-to-end automated testing framework for mobile and web applications. Based on built-in fault tolerance and latency tolerance mechanisms, it solves the instability and waiting problems commonly found in traditional testing. Maestro is based on a declarative syntax,...
OLMo 2 32B is the latest open-source language model from the Allen Institute for AI (Ai2), and a significant achievement in the OLMo 2 series. With 32 billion parameters, it is the first model to surpass GPT-... in multi-skill academic benchmarks.
InternVL is a multimodal large model developed by OpenGVLab at the Shanghai Artificial Intelligence Laboratory, focusing on vision and language tasks. It adopts a ViT-MLP-LLM architecture, consisting of vision modules (such as InternViT) and language modules (such as...
Umi-OCR is a free and open-source offline OCR text recognition software. No internet connection is required; it's ready to use after extraction. It supports text recognition from screenshots, batch images, and scanned PDFs. It can recognize mathematical formulas and QR codes, and can generate two-layer searchable PDFs...
Orpheus TTS is an open-source text-to-speech (TTS) system based on the Llama-3b architecture. Orpheus TTS supports generating natural, emotional, and near-human-level speech, and features zero-sample speech cloning capabilities, requiring no pre-production...
o1-pro is an upgraded version of the o1 series officially released by OpenAI. o1-pro is currently OpenAI's most powerful inference model, with its core advantage lying in its significantly improved computing power, enabling it to better handle complex problems and provide more...
Crack Coder is an open-source, stealthy AI-assisted tool designed specifically for technical interviews. Crack Coder runs in the background, is completely invisible, and cannot be detected by screen recording or monitoring software. Crack Coder provides real-time AI...
YT Navigator is an AI-powered YouTube content search tool that helps users efficiently search and browse YouTube channel content. YT Navigator's natural language search function allows users to quickly locate specific channel videos...
Step-Video-TI2V is an open-source image-to-video generation model developed by StepFun. It boasts 30 billion parameters and can generate videos of up to 102 frames based on text descriptions and image input. ...
Dify-Plus is an enterprise-level enhanced version of Dify, integrating a management center based on gin-vue-admin. Dify-Plus adds features such as user quotas, key quotas, and web public page login to Dify...