ModelsTools

Google's Gemini Omni 1.1 Flash makes AI video generation cheaper and more flexible

Source: The Decoder · Matthias Bastian

Intel Summary

Google has updated its Gemini Omni 1.1 Flash video generation model with enhanced scene extension and a lower-cost preview tier. The model can now analyze up to ten seconds of existing footage for context rather than just the final second, enabling extensions in 10-second increments up to a total duration of 40 seconds. Additionally, a new 360p draft mode is reported to generate previews up to 60 percent faster at one-third of the standard compute cost.

Why It Matters

Context continuity has been a core technical limitation in generative video, where sequential scene extensions frequently suffer from visual drift. Expanding the reference window improves temporal coherence for multi-clip workflows. Concurrently, rapid draft modes reduce experimentation costs for developers and media teams prototyping video assets before committing full rendering budgets.

Part of an ongoing development

Source

Google updated Gemini Omni 1.1 Flash model with enhanced scene extension and 360p draft mode

Google has updated its Gemini Omni 1.1 Flash video generation model with enhanced scene extension and a lower-cost preview tier. The model can now analyze up to ten seconds of existing footage for context rather than just the final second, enabling extensions in 10-second increments up to a total duration of 40 seconds. Claims are as reported; this summary makes no determination about accuracy or significance.

Organizations & Entities