# Kyutai CADRE: Video Autoencoders That Cut World Model Lag Below 230 ms

> **Publication**: aitalky (Independent AI Journalism)
> **Canonical Citation URL**: https://aitalky.vercel.app/news/kyutai-cadre-video-autoencoders-that-cut-world-model-lag-below-230-ms
> **Original Source**: [MarkTechPost](https://www.marktechpost.com/2026/10/23/kyutai-cadre-video-autoencoders-interactive-world-models/)
> **Author**: Michal Sutter (undefined)
> **Published Date**: Friday, October 23, 2026
> **Category**: RESEARCH
> **Tags**: AI, MarkTechPost, research

## Executive Summary
Kyutai released CADRE, a family of 7 open causal diffusion video autoencoders for interactive world models. World models built on CADRE run at 20 fps on 1 H100 and react in under 230 ms



## Full Coverage
Kyutai released CADRE, a family of 7 open causal diffusion video autoencoders for interactive world models. World models built on CADRE run at 20 fps on 1 H100 and react in under 230 ms

---
*Generated for AI agents and LLM citation by aitalky (https://aitalky.vercel.app).*