Alice v1: Distillation-Enhanced Video Generation Surpassing Closed-Source Models

TL;DR

Alice v1 is an open-source 14-billion parameter video generation model that surpasses closed-source systems and its teacher in quality through innovative distillation techniques, achieving high-quality 720p videos efficiently.

Contribution

The paper introduces a novel distillation method with score regularization that improves quality beyond the teacher, along with a targeted synthetic data pipeline and regularization strategies.

Findings

Alice v1 outperforms closed-source models in automated benchmarks.

It generates high-quality 720p videos at 24fps in about 8 seconds.

The model surpasses the teacher's quality, achieving a VBench score of 91.2.

Abstract

Wepresent Alice v1, a 14-billion parameter open-source video generation model that achieves state-of-the-art quality through consistency distillation with score regularization (rCM). Contrary to conventional distillation-which trades quality for speed-we demonstrate that rCM-based distillation can exceed teacher model quality. We attribute this to three mechanisms: (1) the score regularization term acts as a mode-seeking objective that concentrates probability mass on high-quality outputs rather than covering the full teacher distribution, (2) our targeted synthetic data pipeline with hard example mining provides training signal specifically for failure modes (physics, hands, faces) that the teacher handles inconsistently, and (3) consistency enforcement acts as implicit regularization, eliminating "lucky path" dependence on specific noise samples. Alice v1 generates 5-second 720p videos at 24fps in 4 denoising steps (~8 seconds on H100), a 7x speedup over the 50-step teacher while improving VBench score from 84.0 (Wan2.2) to 91.2. This surpasses both the teacher and closed-source systems including Veo3 (~90) and Sora2 (~88) on automated benchmarks, with competitive results in human preference studies. We release all model weights, training code, synthetic data pipelines, and evaluation scripts to advance open research in video generation.