🔍 Intelligence Synthesis JP/EN
Ethics · Regulation · Technology — Pharma Practice Notes

AI Tech Flash ── 2026-11-06

Google makes Gemini Omni 1.2 Flash video model generally available

Up to 15-second clips at $0.07 per second of 720p on Google Cloud.

What changes

According to Google Cloud's announcement, the model can generate and edit video for up to 15 seconds. The previous limit was 10 seconds.

Google Cloud says the model follows image and video references more closely. It says products, characters and brand details stay consistent from shot to shot. The company also says users can refine the same clip in conversation instead of starting over.

Why it matters

In Google Cloud's announcement, 720p video costs $0.07 per second. That is 30% less than the earlier Gemini Omni 1.1 Flash.

MarkTechPost calculates that a 15-second 720p clip costs about $1.05. For businesses that make product videos and ads in volume, this makes costs easier to predict.

However, MarkTechPost reports that no benchmark scores (standardized comparison tests) have been published for version 1.2. The performance claims come from Google's own description.

What it is

Gemini Omni 1.2 Flash is an AI model that generates and edits video. According to MarkTechPost, Google DeepMind built it.

MarkTechPost says it accepts text, images, short video clips and audio as input. It outputs video with audio. The outlet also says the weights are closed, so the model runs only on Google-hosted services.

Who and where

According to Google Cloud, the model is now generally available on Google Cloud. Business users can use it in Google Vids. It is coming soon to Gemini for business. Developers can use it through the API.

MarkTechPost reports it is also available through the Gemini API, and that Google Vids users can generate 15-second clips at no additional cost. Google Cloud notes that audio references require allowlisting.

Why it was built

Google Cloud says the model is built for everyday videos that businesses make in volume. Its examples are product videos, virtual try-ons and ads.

The company describes the goals as fast, cost-effective and reliable video creation. Partner Photoroom says 15-second video was a top user request.

How it works

According to Google Cloud, users first make drafts at 360p or 720p. They then upscale only the best ones to 1080p or 4K. Upscaling means producing the video again at a higher resolution.

MarkTechPost says edits reference the previous interaction by an identifier. The earlier video therefore does not need to be uploaded again. Users can swap a product, background or lighting one step at a time.

Neither source gives details of the model's internal design or training.

Key points

  1. Google Cloud says clips can run up to 15 seconds, up from 10.
  2. In Google Cloud's announcement, 720p video costs $0.07 per second, 30% less than the prior version.
  3. MarkTechPost reports that no benchmark scores for version 1.2 have been published.

Conclusion

Google Cloud has made Gemini Omni 1.2 Flash generally available. The main changes are 15-second clips, a 30% price cut and conversational editing.

The performance claims come from Google. MarkTechPost reports that no comparison tests for version 1.2 are published, so real-world quality still depends on user feedback.

Quotes and sources

“Today, Gemini Omni 1.2 Flash is generally available on Google Cloud.”
── Google Cloud, 2026-11-05
“A lower price: $0.07 per second of 720p video, 30% less than Gemini Omni 1.1 Flash, with similar savings at other resolutions.”
── Google Cloud, 2026-11-05
“Google has not published benchmark scores for version 1.2.”
── MarkTechPost, 2026-11-06

Source articles

Written by AI (Claude) from the cited articles, then checked by Jev (TypeSafe's judgement model), a Gemini fact-check and a verbatim check of every quote.