minssam.
Published on

Gemini Omni 1.2 Flash β€” 15-Second AI Video and Conversational Editing

Tell the AI "make it a bit brighter" and the video changes.

On November 5, 2026, Google released Gemini Omni 1.2 Flash. Compared to its predecessor, Omni 1.1 Flash, two things changed: videos got longer, and you can now edit them through conversation.

What Changed

1. 15-Second Video β€” Up 50% from 10 Seconds

Gemini Omni 1.1 Flash generated videos up to 10 seconds. Omni 1.2 Flash extends that to 15 seconds. Five seconds may not sound like much, but in short-form content it is a meaningful unit. YouTube Shorts standard is up to 60 seconds, and Instagram Reels short clips cluster around the 15-second mark.

Practically, 15 seconds allows a more complete scene. At 10 seconds, an entrance and an action were rushed. At 15 seconds, a brief setup and a landing moment both fit naturally.

2. Multi-Turn Conversational Editing β€” Keep Talking to Refine

The more significant change in this update is how editing works. Omni 1.2 Flash lets you refine a generated video through ongoing conversation rather than starting from scratch each time.

The old approach was to revise the prompt and regenerate an entirely new video. That meant the parts you liked could disappear along with the parts you wanted to change.

Conversational editing works differently:

  1. Generate a video with an initial prompt.
  2. Say "make the background darker."
  3. Receive a version with only the background adjusted.
  4. Say "move the subject a bit more to the left."
  5. Receive a version with the position corrected.

You converge toward the result you want through dialogue.

Comparison with Previous Version

FeatureOmni 1.1 FlashOmni 1.2 Flash
Max video length10 seconds15 seconds
Conversational editingNot availableAvailable (multi-turn)
Release dateEarly 2026November 5, 2026

What Is the Gemini Omni Series?

Gemini Omni is Google''s family of multimodal models that accept text, images, audio, and video as input and output video. While earlier Veo models generated video from text or image input, Omni can combine multiple input types.

"Flash" in the model name refers to the speed- and cost-optimized variant in the Gemini lineup. It is designed for fast iteration rather than maximum quality.

Who Can Use It?

Gemini Omni 1.2 Flash is accessible through the Gemini app. Limited use is available on the free plan. Greater usage and higher-resolution options are offered on paid plans (Plus, Pro, Ultra).

Why Conversational Editing Matters

The practical limit of AI video tools has always been "you rarely get what you want on the first try." Each fine-tuned adjustment required a full regeneration, wasting time and generation credits.

Conversational editing brings the experience closer to actual creative work. Refining a video through dialogue is conceptually similar to adjusting layers in a photo editor or trimming clips on a timeline β€” except you do it in plain language.

Edtech Perspective β€” Student Creative Projects

Short "introduction video" assignments are already common in schools. But the tools have remained frustrating because editing a generated video was nearly as much work as starting over.

Omni 1.2 Flash''s conversational editing lets a student say "change this part" and iteratively improve their work without breaking the creative flow. Creation and revision merge into a single conversation.

Tips for Use

  • Write the first prompt as clearly as possible. Use the multi-turn editing to dial in the details afterward β€” that division of labor makes the process much more efficient.
  • When editing conversationally, change one thing at a time rather than stacking multiple requests in one message. Results are more predictable that way.
  • For 15-second videos, include "three-part structure: setup, action, closing" in the prompt for a more self-contained result.
  • Specify "9:16 portrait orientation" for vertical social media content.

Sources:

Gemini Omni 1.2 Flash β€” 15-Second AI Video and Conversational Editing | MINSSAM.COM