[ICLR 2026] Data Pipeline, Models, and Benchmark for Omni-Captioner.

利用AI大模型，一键生成高清短视频 Generate short videos with one click using AI LLM.

Official Code of NAVA: Native Audio-Visual Alignment for Generation.

​​Unlimited-length talking video generation​​ that supports image-to-video and video-to-video generation

PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusion

Code for "OmniNFT: Modality-wise Omni Diffusion Reinforcement for Joint Audio-Video Generation"

Lens is a 3.8B-parameter text-to-image diffusion model that achieves quality competitive with and in several cases surpassing models like FLUX and SD3, while requiring significantly less training c…

VEFX-Bench: A Holistic Benchmark for Generic Video Editing and Visual Effects

Unified Codebase for Advanced World Models.

Inference script for Oasis 500M

Matrix-Game 3.0: Real-Time and Streaming Interactive World Model with Long-Horizon Memory

Official implementation of Tuna-2: Pixel Embeddings Beat Vision Encoders for Unified Understanding and Generation

open source code for Tencent tFold

ASID-Caption: Attribute-Structured and Quality-Verified Audiovisual Instruction Dataset and Training Pipeline for Fine-Grained Video Understanding.

Implementation of D4RT, Efficiently Reconstructing Dynamic Scenes, from Deepmind

A Distributed Attention Towards Linear Scalability for Ultra-Long Context, Heterogeneous Data Training

A framework for efficient model inference with omni-modality models

Project Lyra: Open Generative 3D World Models

Gen-Searcher: Reinforcing Agentic Search for Image Generation

[ICML'26] Code and website for Self-Flow: Self-Supervised Flow Matching for Scalable Multi-Modal Synthesis

AI agents running research on single-GPU nanochat training automatically

Out of Sight but Not Out of Mind: Hybrid Memory for Dynamic Video World Models

A feed-forward 3D foundation model for reconstructing scenes from streaming data

[NeurIPS 2025 Oral] Representation Entanglement for Generation: Training Diffusion Transformers Is Much Easier Than You Think

[ICLR'25 Oral] Representation Alignment for Generation: Training Diffusion Transformers Is Easier Than You Think

JoyAI-Image is the unified multimodal foundation model for image understanding, text-to-image generation, and instruction-guided image editing.