Two months after the launch of Gemini Omni Flash, Google releases version 1.1 with a concrete leap in duration and creative control: scenes can now be extended up to 40 seconds in 10-second increments. The update, available on Google AI Studio and documented in the official Google blog post, transforms the model from a rapid generation tool into a platform for more structured workflows.

Add AlexTech.ai asPreferred Source on Google

Scene extension and keyframe control

The main technical improvement is the analysis of up to 10 seconds of existing video for extensions, instead of just the last second. This allows visually more consistent results when lengthening a clip. Developers can also upload up to 3 seconds of external footage as a style reference to transfer characters or motion patterns, and set start and end frames to create camera movements between keyframes. This is a step toward non-linear control of video narrative, previously the domain of traditional editing software.

360p draft at one-third cost and 4K upscaling

Per-second pricing introduces a resolution hierarchy: $0.03 at 360p (draft), $0.10 at 720p, $0.15 at 1080p, and $0.30 at 4K. The 360p draft mode is up to 60% faster and costs one-third compared to 720p, making rapid iterations economically sustainable. Videos generated in low resolution can be upscaled to 1080p or 4K, offering a compromise between prototyping speed and final quality. Compared to the previous version, 720p remains at parity ($0.10/s), but the absence of a 360p tier in the original model makes the new tier a competitive advantage for high-volume generation.

Context: from Gemini Omni Flash to 1.1

As reported in Google accelerates with Nano Banana 2 Lite and Gemini Omni Flash, the original model was positioned as a low-cost, developer-first tool. Version 1.1 maintains this philosophy but adds the four missing capabilities: scene extension, keyframe control, cheap drafts, and 4K output. Integration with Google Photos shows how the platform is maturing from a pure API into a complete creative ecosystem, where the video model fits into conversational editing pipelines.

Implications for AI video workflows

The combination of extended duration, keyframe control, and tiered pricing positions Gemini Omni 1.1 Flash as a serious alternative for short- to medium-form content production, not just demo clips. The ability to iterate on low-cost drafts before final rendering reduces creative risk and failure costs. For developers integrating video generation into applications, start/end frame control opens interactivity scenarios impossible with previous models, where output was deterministic and unmodifiable post-generation.