TurboDiffusion: 100–200× Acceleration for Video Diffusion Models

Official Implementation of "MemFlow: Flowing Adaptive Memory for Consistent and Efficient Long Video Narratives"

[Siggraph Asia 25] SS4D: Native 4D Generative Model via Structured Spacetime Latents

The paper list of "Memory in the Age of AI Agents: A Survey"

WorldPlay: Interactive World Modeling with Real-Time Latency and Geometric Consistency

V-RGBX: Video Editing with Accurate Controls over Intrinsic Properties

"E-RayZer: Self-supervised 3D Reconstruction as Spatial Visual Pre-training" official implementation.

Visionary: The World Model Carrier Built on WebGPU-Powered Gaussian Splatting Platform

[NeurIPS 2025] The official repository of "Sekai: A Video Dataset towards World Exploration"

[NeurIPS 2025] Wan-Move: Motion-controllable Video Generation via Latent Trajectory Guidance

Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation

[ArXiv 25] Stable Video Infinity: Infinite-Length Video Generation with Error Recycling

ViSAudio: End-to-End Video-Driven Binaural Spatial Audio Generation

Official inference repo for FLUX.2 models

SAM 3D Objects

Cambrian-S: Towards Spatial Supersensing in Video

A list of open source games.

Depth Anything 3

[ICML2025, NeurIPS2025 Spotlight] Sparse VideoGen 1 & 2: Accelerating Video Diffusion Transformers with Sparse Attention

ChronoEdit: Towards Temporal Reasoning for Image Editing and World Simulation

We introduce 'Thinking with Video', a new paradigm leveraging video generation for multimodal reasoning. Our VideoThinkBench shows that Sora-2 surpasses GPT5 by 10% on eyeballing puzzles and reache…

Native Multimodal Models are World Learners

Krea Realtime 14B. An open-source realtime AI video model.

Thinking with Camera: A Unified Multimodal Model for Camera-Centric Understanding and Generation

This is the official implementation for Human-Agent Collaborative Paper-to-Page Crafting for Under $0.1.

[Preprint 2025] Ditto: Scaling Instruction-Based Video Editing with a High-Quality Synthetic Dataset

🚀🚀 「大模型」2小时完全从0训练26M的小参数GPT！🌏 Train a 26M-parameter GPT from scratch in just 2h!

Infinite-Forcing: Towards Infinite-Long Video Generation