Visko Orbis 1.0: A Live Model for Real-Time Interactive Long Video Generation

2026-07-29

AI Summary

Existing video generation models struggle to simultaneously achieve real-time interactivity, dynamic prompt editing, and visual consistency over extended durations. This work proposes the first long-form video generation system capable of hour-scale real-time interaction, enabling users to modify multilingual prompts on-the-fly during generation—with immediate reflection in the output—across text-to-video, image-to-video, and video continuation modes. The system ensures cross-segment consistency in subject, scene, and style through a bounded multi-scale memory mechanism, while a distillation-driven chunked streaming generator, a streaming super-resolution module, and an optimized GPU serving engine jointly enable high-quality real-time synthesis at 4K resolution and 24 FPS. Experiments demonstrate that the proposed system significantly outperforms existing approaches in long-video benchmarks, achieving state-of-the-art results in both overall user preference and temporal stability.

Abstract

We present Visko Orbis 1.0, a Live Model for real-time, interactive long-video generation. Users can change the prompt at any moment during generation, and the update becomes visible in real time. Visko Orbis 1.0 supports long-form text-to-video, image-to-video, and video continuation, with multilingual prompts and prompt switching while generation is in progress. A bounded multi-scale memory preserves subjects, scenes, and style across chunks, sustaining hour-scale rollouts without evident quality or color drift. Built on a distilled chunk-wise streaming generator and a streaming video upscaler, Visko Orbis 1.0 delivers real-time 4K video generation at 24 FPS using an optimized GPU serving engine. In long-form Arena comparisons, Visko Orbis 1.0 obtains the highest overall-preference and temporal-stability ratings among state-of-the-art real-time interactive video-generation systems.

Problem

• real-time video generation
• interactive video generation
• long-form video
• temporal stability
• prompt switching

Innovation

• real-time video generation
• interactive prompting
• long-form video synthesis
• streaming generative model
• temporal consistency