Step-3
Step-3 is an open-weight large multimodal mixture of experts (MoE) model from StepFun, the Shanghai-based Chinese artificial intelligence startup also known as Jieyue Xingchen.
Explore Multimodal AI through related topics and the articles other pages reference most.
Articles that also belong to these categories. Counts cover all of Multimodal AI.
Showing 121-138 of 138 articles
Step-3 is an open-weight large multimodal mixture of experts (MoE) model from StepFun, the Shanghai-based Chinese artificial intelligence startup also known as Jieyue Xingchen.
T-Rex: Tactile-Reactive Dexterous Manipulation is a 2026 robot learning research system for contact-rich, bimanual manipulation.
Griffin is Tavus's full-duplex audiovisual model for face-to-face conversation, announced on October 1, 2026. Tavus calls it a Human Interaction Model; Griffin-Lite is the research preview.
Tencent Cloud WAND is Tencent Cloud's brand for AI capabilities used in audio and video services. Its WAND-Vega family includes image and video generation, multimodal understanding and audio processing.
Text-to-image models are generative artificial intelligence systems that synthesize a new image from a natural-language description, called a prompt.
UI-TARS is a native graphical user interface (GUI) agent model developed by ByteDance through its Seed research team.
Video-MME (Video Multi-Modal Evaluation) is a benchmark for testing how well multimodal large language models (MLLMs) understand video, built from 900 manually selected videos totaling 254 hours and 2,700…
A vision encoder is the neural network component that turns an image into a sequence of numerical vectors that other models can consume.
Vision-language models (VLMs) are artificial intelligence models that learn relationships between visual data and natural language.
A vision-language-action model (VLA model, or VLA) is a machine-learning model for robotics that uses visual observations and a natural-language instruction to generate actions a robot can execute.
Visual question answering models are AI systems that take an image and a natural language question about that image and return a natural language answer.
Visual question answering (VQA) is the task of producing a natural-language answer to a natural-language question about an image.
WeMM-Embedding (WeChat Multi-Modal Embedding) is a family of open-weight universal multimodal embedding models built by the WeChat Vision team at Tencent.
WeVisOne-VL (WeChat Vision All-in-One) is a vision-language model family developed by the WeChat Vision team at Tencent.
ZeroBench is a visual reasoning benchmark built to be effectively impossible for current frontier large multimodal models, which score 0.0% on its main questions.
d3-turbo is a multimodal decision model in the Decision 3.0 family from the vLLM Semantic Router team. It produces structured decisions and probabilities rather than free-form generated text.
data2vec is a self-supervised learning framework from Meta AI (then Facebook AI Research) that applies the same training method to three different input types: speech, computer vision, and text.
dots3-note Preview is an open-weight multimodal model and large language model developed by dots studio, an AI team at Xiaohongshu.