Google just made AI video generation longer, cheaper, and considerably more interesting.
Meet Gemini Omni 1.2 Flash, the latest update to Google's AI video generation model.
And honestly, the most exciting feature isn't the longer clips or the lower price.
It's the ability to edit your videos through an ongoing conversation.
Imagine generating a cinematic scene and then simply telling the model:
"Change the background to a forest."
"Make the lighting more dramatic."
"Replace the product with this one."
"Now create a vertical version for YouTube Shorts."
And the AI continues working on the same video without requiring you to describe the entire scene from scratch.
That's the real upgrade.
Here's what's new in Gemini Omni 1.2 Flash:
15-second videos: Up from the previous 10-second limit.
30% cheaper generation: Approximately $0.07 per second at 720p.
Conversational editing: Change objects, backgrounds, lighting, weather, and aspect ratios through follow-up prompts.
Better character consistency: Google claims improved preservation of faces, clothing, and product details.
More natural dialogue: Improved lip synchronization and conversations involving multiple characters.
Smoother camera movement: Fewer unwanted cuts during complex camera movements, including FPV flythroughs.
Reference-based generation: Use images and other supported visual references to guide the result.
Multiple resolutions: Generate at 360p or 720p, with higher-resolution output options up to 4K.
And the pricing is getting interesting.
At 720p, Google's new output price is approximately:
10-second video: $0.70
15-second video: $1.05
60 seconds of generated footage: $4.20
Those figures are based on video output pricing. Input tokens, additional edits, and other usage can increase the total cost.
But here's what I find particularly useful.
Gemini Omni 1.2 Flash supports multi-turn editing through Google's Interactions API.
Instead of treating every generation as a completely separate request, developers can preserve the conversation context and continue refining their video.
That's an important shift.
Because real creative production rarely works like this:
Write one perfect prompt. Generate one perfect video. Done.
It usually looks more like:
Generate. Review. Change the lighting. Fix the character. Adjust the camera. Change the background. Try again.
And that's exactly the kind of workflow AI video tools need to support.
Google also claims that Omni 1.2 Flash is better at maintaining visual consistency across multiple shots and complex motion.
But there are still limitations.
Fast sports movements, complicated physics, and rapid changes in the storyline remain challenging.
And Google's quality improvements are currently based primarily on its own demonstrations, rather than extensive independent comparisons.
Availability:
Gemini Omni 1.2 Flash is now generally available through the paid Gemini API and Google Cloud's Vertex AI.
The model identifier is:
gemini-omni-1.2-flash
And if you're already using the previous version, there's an important date to remember.
Starting December 3, 2026, API requests to gemini-omni-1.1-flash will automatically route to version 1.2 at the new pricing.
Regional availability and feature restrictions still apply.
And here's the bigger picture.
We've spent years getting better and better at generating short AI videos.
But the next major improvement might not be another jump in visual quality.
It might be the ability to actually direct and revise those videos without starting over every time.
Less prompt lottery.
More creative control.
And at $0.07 per second, experimenting is becoming considerably more affordable.
The future of AI filmmaking might look less like generating clips and more like having an ongoing conversation with your editor.
Google's official announcement:
https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemini-omni-1-2-flash/
Gemini API pricing:
https://ai.google.dev/gemini-api/docs/pricing#gemini-omni-1.2-flash
Gemini Interactions API:
https://ai.google.dev/gemini-api/docs/interactions-overview
Supported countries and regions:
https://ai.google.dev/gemini-api/docs/available-regions
#GeminiOmni #GeminiOmni12 #GoogleAI #AIVideo #AIVideoEditing #GenerativeAI #AIFilmmaking
YouTube / ABVX