Qwen3-Omni is the industry's first native end-to-end multimodal AI model launched by Alibaba's Tongyi team, capable of seamlessly processing multimodal data such as text, images, audio, and video. The model achieved a score of 22 out of 36 in audio and video benchmark tests...
Qwen3-TTS-Flash is a flagship speech synthesis model launched by Alitongyi, supporting multiple timbres, languages, and dialects. The model boasts excellent stability in both Chinese and English pronunciation, outstanding multilingual performance, and highly expressive human-like timbres...
Qianfan-VL is a large-scale visual understanding model launched by Baidu AI Cloud Qianfan, designed for enterprise-level multimodal applications. The model is available in 3B, 8B, and 70B sizes, boasting excellent versatility and targeted capabilities for vertical applications such as OCR and education...
Doubao-Seed-Translation is an advanced multilingual translation model launched by the ByteDance team, supporting mutual translation between 28 languages, covering a variety of commonly used languages such as Chinese, English, Japanese, Korean, German, and French.
presentation-ai is an open-source AI presentation generation tool developed by the ALLWEONE team. It can automatically generate complete PPT outlines and beautiful slides based on user-input topics, supporting multiple languages, themes, and customizable formats...
Qwen3Guard is the first guardrail model in the Qwen family specifically designed for security protection, launched by the Alibaba Tongyi team. It's built on the robust Qwen3 architecture. The model has been specifically fine-tuned for security classification tasks, enabling efficient...
Qwen3-Max is a massively multi-scale language model from the Qwen team, with over 1T parameters and 36T tokens used in pre-training. Qwen3-Max is currently the largest and most powerful model in the Qwen series, possessing outstanding...
Qwen3-VL is the most powerful visual language model in the Qwen series launched by Alibaba Tongyi, possessing outstanding multimodal capabilities. The model can understand plain text, images, and videos, and supports long context, spatial awareness, code generation, and other functions...
Qwen3-LiveTranslate is a multilingual real-time audio and video simultaneous interpretation model developed by the Ali Tongyi team, based on a large language model. The model supports translation of 18 languages and multiple dialects, and features visual enhancement technology that can combine lip-syncing...
Wan2.5 is Alibaba's latest multimodal generative model, a preview version (Wan2.5-Preview) encompassing four main functions: text-to-video, image-to-video, text-to-image, and image editing...
MiroFlow v0.2 is an open-source research agent framework launched by MiroMind. It can upgrade any large language model (LLM) to the level of OpenAI deep research capabilities, focusing on reliably completing complex tool usage tasks.
LucaVirus is a unified nucleic acid-protein language model developed by Alibaba Cloud LucaGroup, specifically designed for viruses. Trained on 25.4 billion nucleotide and amino acid tags, it covers almost all known viruses. The model can learn...
Audio2Face is an AI-powered facial animation generation model from NVIDIA that generates realistic facial animations from audio input. The model analyzes phonemes and intonation in audio to drive lip movements and facial expressions, achieving precise...
CWM (Code World Model) is a 32 billion parameter code world model open-sourced by Meta. It is the world's first language model to systematically introduce the world model into code generation, supporting code generation and understanding.
Neovate Code is an open-source intelligent programming assistant from Ant Group, focused on improving development efficiency. The tool deeply understands codebases, follows existing coding practices, and accurately completes tasks such as code generation, bug fixing, and refactoring.
Vidu Q2 is a new generation of image-generated video model launched by Shengshu Technology, realizing a leap from "video generation" to "acting generation" in AI video generation technology, marking a new era in AI video development that pursues "lifelikeness." Vidu Q2 can generate...
AudioFly is an open-source AI model from iFlytek for generating audio effects from text. The model uses a latent diffusion model architecture, has 1 billion parameters, and utilizes a large number of open datasets (such as AudioSet, AudioCaps, and TUT) as well as internal proprietary data...
SimpleFold is a lightweight AI model for predicting protein folding, developed by Apple. Based on flow matching technology, it skips complex modules such as multiple sequence alignment (MSA) and directly generates protein folding predictions from random noise...
FLM-Audio is a native full-duplex audio dialogue model jointly released by the Beijing Academy of Artificial Intelligence, Spin Matrix, and Nanyang Technological University in Singapore. It supports both Chinese and English. Employing a native full-duplex architecture, it can...
Spark Chemistry-X1-13B is a large-scale language model for chemistry, open-sourced by iFlytek. Based on the iFlytek Spark X1-0420 model, it has been fine-tuned across various chemistry task datasets, demonstrating excellent capabilities in solving complex chemical problems...
GDPval is a new evaluation framework launched by OpenAI to measure the performance of AI models on real-world economic value tasks. GDPval selects 44 occupations from the 9 industries that contribute the most to US GDP and designs 1320...
Hunyuan3D-Omni is a 3D asset generation framework proposed by Tencent's Hunyuan3D team. It achieves precise 3D model generation through various control signals. Based on the Hunyuan3D 2.1 architecture, it introduces a unified control programming...
Hunyuan 3D-Part is a 3D generative model launched by Tencent, composed of P3-SAM and X-Part. It achieves high-precision, controllable component-based 3D generation for the first time, supporting the automatic generation of 50+ components. Users can first use Hunyuan 3D 2.5 or 3.0 to generate...
StableAvatar is an innovative audio-driven virtual avatar video generation model developed by Fudan University, Microsoft Research Asia, and others. The model utilizes an end-to-end video diffusion transformer, combined with a time-step-aware audio adapter and native audio...