Paper 2605.30263
minWM: A Full-Stack Open-Source Framework for Real-Time Interactive Video World Models
- Published
- May 2026
- Research lab
- Independent
- Citations
- 6
- GitHub
- 766 stars
01 In brief
Summary
minWM is a full-stack open-source framework for converting bidirectional text-to-video (T2V) or text-and-image-to-video (TI2V) diffusion foundation models into camera-controllable, few-step autoregressive (AR) world models for real-time interaction.
The pipeline has two phases: first, fine-tuning the bidirectional model with camera control via PRoPE (projective relative positional encoding); second, applying Causal Forcing or Causal Forcing++ distillation, which includes AR diffusion training, causal ODE or causal consistency distillation initialization, and asymmetric DMD post-training.
The framework is instantiated on Wan2.1-T2V-1.3B and HY1.5-TI2V-8B, and supports adapting existing world models like HY-WorldPlay.
Experiments show few-step AR models reduce first-frame latency by 223.75x (HY1.5) and 236.64x (Wan2.1) over multi-step bidirectional baselines, while preserving camera controllability.
Ablations reveal that ground-truth camera trajectories (from 3D reconstruction or WorldPlay) are crucial, controllability emerges after ~5K training steps and is strong at 8K steps, and a minimum batch size of 16 is needed for stable training.
02 From the paper
Abstract
Recent video diffusion foundation models have achieved remarkable progress in high-quality video generation, yet turning them into real-time interactive video world models remains challenging. Interactive world models require controllable, causal, and low-latency rollout, which in practice demands a full pipeline spanning data construction, controllable fine-tuning, autoregressive training, few-step distillation, and streaming inference. In this work, we present minWM, a full-stack open-source framework for building real-time interactive video world models. minWM provides an end-to-end pipeline that converts existing bidirectional T2V/TI2V video foundation models into camera-controllable few-step autoregressive world models. Specifically, minWM first fine-tunes a bidirectional video diffusion model with camera control, and then applies the Causal Forcing / Causal Forcing++ pipeline, including AR diffusion training, causal ODE or causal consistency distillation, and asymmetric DMD, to distill it into a few-step autoregressive generator for low-latency rollout. The framework is modular and architecture-extensible: we instantiate it on representative open backbones, including Wan2.1-T2V-1.3B and HY1.5-TI2V-8B, covering both cross-attention-based condition injection and MMDiT-style architectures. minWM also supports adapting existing video world models, such as HY-WorldPlay, to new data distributions, training recipes, and latency targets. Beyond releasing runnable scripts, checkpoints, documentation, and inference code, we provide practical ablations on camera trajectory quality, controllability training steps, and minimal batch-size requirements. We hope minWM serves as a reproducible and extensible recipe for building and adapting real-time interactive video world models. Project Page: [https://github.com/shengshu-ai/minWM](https://github.com/shengshu-ai/minWM)