Paper page - AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation

Source: https://huggingface.co/papers/2605.13724

Abstract

AnyFlow introduces a novel any-step video diffusion distillation framework that improves upon consistency distillation by optimizing full ODE sampling trajectories through flow-map transition learning and backward simulation techniques.

Few-stepvideo generationhas been significantly advanced byconsistency distillation. However, the performance of consistency-distilled models often degrades as more sampling steps are allocated at test time, limiting their effectiveness for any-step video diffusion. This limitation arises becauseconsistency distillationreplaces the original probability-flow ODE trajectory with a consistency-sampling trajectory, weakening the desirable test-time scaling behavior ofODE sampling. To address this limitation, we introduce AnyFlow, the first any-step video diffusion distillation framework based onflow maps. Instead of distilling a model for only a few fixed sampling steps, AnyFlow optimizes the fullODE samplingtrajectory. To this end, we shift the distillation target from endpoint consistency mapping (z_{t}rightarrow z_{0}) to flow-map transition learning (z_{t}rightarrow z_{r}) over arbitrary time intervals. We further propose Flow Map Backward Simulation, which decomposes a fullEuler rolloutinto shortcut flow-map transitions, enabling efficienton-policy distillationthat reduces test-time errors (i.e.,discretization errorin few-step sampling andexposure biasincausal generation). Extensive experiments across both bidirectional and causal architectures, at scales ranging from 1.3B to 14B parameters, demonstrate that AnyFlow achieves performance matches or surpasses consistency-based counterparts in the few-step regime, while scaling with sampling step budgets.

View arXiv pageView PDFProject pageGitHubAdd to collection

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2605.13724 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2605.13724 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2605.13724 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.