DCLM-7B is a small, open-source model with 7 billion parameters, developed in collaboration with a research team at Apple. Its performance surpasses Mistral-7B and approaches that of Llama 3 and Gemma. Recently, Apple, as a research institution for the DataComp-LM (DCLM) project...
LivePortrait is an open-source portrait animation generation framework launched by Kuaishou. It focuses on efficiently and controllably transferring the expressions and poses that drive videos to static or dynamic portraits, creating expressive videos. This technology uses implicit...
Llama 3.1 is Meta's latest open-source AI model, comprising three versions: 8B, 70B, and 405B. Version 405B, with its 405 billion parameters, is one of the largest open-source models currently available. Llama 3.1 supports 128K context lengths...
The GPT-4o Long Output model is OpenAI's latest ultra-long output AI model, supporting text output of up to 64k tokens, equivalent to approximately 200 pages of a novel. Compared to the original GPT-4o model, its output capability is improved by 16 times, but...
MindSearch is an open-source AI search framework developed by a joint research team at the Shanghai Artificial Intelligence Laboratory. It combines large-scale information gathering and processing capabilities. Utilizing the InternLM2.5 7B dialogue model, MindSearch can retrieve data from...
Diffree is an AI texture mapping framework developed by OpenGVLab. It automatically finds suitable locations in an image and adds new objects based on user-defined text descriptions. Through intelligent mask prediction, it determines the location of new objects without requiring manual drawing...
FLUX.1 is an open-source AI image generation model developed by the founding team of Stable Diffusion. With 12B parameters, it is the largest text-to-image generation model to date. It includes three variants: the top-performance FLUX.1 [pro], an open-source non-commercial...
Stable Fast 3D (SF3D) is an innovative 3D mesh reconstruction technology developed by Stability AI, capable of generating high-quality 3D models from a single image in 0.5 seconds. Stable Fast 3D utilizes advanced Transformer networks...
Tora is an AI video generation framework launched by Alibaba. Based on trajectory-guided diffusion transformer (DiT) technology, it fuses text, visual information, and trajectory conditions to generate high-quality video content that conforms to the dynamics of the physical world. Tora consists of trajectory...
Whisper-Medusa is an open-source AI speech recognition model launched by aiOla. Combining OpenAI's Whisper technology with aiOla's innovations, Whisper-Medusa introduces a multi-head attention mechanism, enabling parallel processing and significantly improving...
FoleyCrafter is an AI video dubbing framework jointly developed by the Shanghai Artificial Intelligence Laboratory and the Chinese University of Hong Kong, Shenzhen. FoleyCrafter can automatically detect actions in a video and add appropriate sound effects. Whether it's...
MinerU is an open-source intelligent data extraction tool developed by the OpenDataLab team at the Shanghai Artificial Intelligence Laboratory. It focuses on the efficient parsing and extraction of complex PDF documents. MinerU can extract data from multi-modal PDF documents containing images, formulas, tables, and other elements...
CogVideoX is the latest open-source AI video generation model from Zhipu AI, sharing the same origins as Zhipu AI's commercial product "Qingying". CogVideoX supports English prompts and can generate 6-second videos at 8 frames per second with a resolution of 720*480. ...
PhotoMaker V2 is an AI image generation framework launched by Tencent, capable of generating realistic portrait photos in a very short time. Compared to the first generation, version V2 has significantly improved the consistency and controllability of characters, allowing users to...
SEED-Story is a multimodal story generation model developed by Tencent in collaboration with the Hong Kong University of Science and Technology and the Chinese University of Hong Kong. Based on a multimodal large language model (MLLM), it can predict text and visual tokens, and generate stories through a visual de-tokenizer...
FlashFace is a high-fidelity AI portrait tool developed by Alibaba in collaboration with the University of Hong Kong. It can quickly generate personalized, high-fidelity portrait photos based on user-provided facial images and text prompts. FlashFace features high-fidelity identity verification...
EasyAnimate is an AI video generation tool launched by Alibaba, supporting both text-based and image-based video generation. Users can upload images as the start and end frames of their videos, enabling more flexible video editing. EasyAnimate...
EmoTalk3D is a 3D digital human framework jointly developed by Huawei Noah's Ark Lab, Nanjing University, and Fudan University. The core technology lies in its ability to synthesize 3D talking avatars with rich emotional expression. EmoTalk3D can capture and reproduce...
SAM 2 (Segment Anything Model 2) is an AI object segmentation model from Meta, focusing on real-time image and video object segmentation. It features zero-shot generalization capability, accurately segments unknown objects, and utilizes a unified architecture...
Wordware is an integrated development environment (IDE) that allows anyone to build complex AI agents and applications using everyday language. Users don't need programming knowledge to quickly develop customized applications using pre-built tools and models...
Qwen2-Math is an open-source AI model for solving mathematical problems, developed by Alibaba's Tongyi Qianwen platform and built upon the Qwen2 language model. It is specifically designed for solving complex mathematical problems. After pre-training on a dedicated mathematical corpus and fine-tuning with instructions, it demonstrates...
Aide is a free and open-source AI programming plugin that features code comments, language conversion, smart pasting, batch processing, variable naming, and custom commands to help developers quickly understand, write, and refactor code. Aide supports...
MaxKB is an open-source AI knowledge base question-answering system launched by Feizhi Cloud. It offers out-of-the-box functionality, supporting document upload, online document crawling, automatic text splitting, and vectorization. Users can quickly embed MaxKB into third-party business systems...
ExAvatar is a 3D digital avatar generation model jointly developed by DGIST and Meta's Codec Avatars Lab. It clones a human image from a short video and converts it into a 3D digital avatar. It supports mobile phone scanning to quickly capture full-body movements and expressions...