InfiniteYou | ByteDance Intelligent Creation

Project landing page for WeDLM

LeVo: High-Quality Song Generation with Multi-Preference Alignment

GITHUB HUGGING FACE MODELSCOPE DISCORD Today, we’re announcing Qwen3-Coder, our most agentic code model to date. Qwen3-Coder is available in multiple sizes, but we’re excited to introduce its most powerful variant first: Qwen3-Coder-480B-A35B-Instruct — a 480B-parameter Mixture-of-Experts model with 35B active parameters which supports the context length of 256K tokens natively and 1M tokens with extrapolation methods, offering exceptional performance in both coding and agentic tasks. Qwen3-Coder-480B-A35B-Instruct sets new state-of-the-art results among open models on Agentic Coding, Agentic Browser-Use, and Agentic Tool-Use, comparable to Claude Sonnet 4.

Kimi K2 Thinking, Moonshot's best open-source thinking model.

Abstract page for arXiv paper 2601.13976: FantasyVLN: Unified Multimodal Chain-of-Thought Reasoning for Vision-Language Navigation

Space-Time Text-to-Video diffusion model by Google Research.

Abstract page for arXiv paper 2601.14251: LightOnOCR: A 1B End-to-End Multilingual Vision-Language Model for State-of-the-Art OCR

Abstract page for arXiv paper 2602.04804: OmniSIFT: Modality-Asymmetric Token Compression for Efficient Omni-modal Large Language Models