Skip to content

Gemini Omni 1.2 Is Now in Phosphene

Gemini Omni 1.2 makes video with native sound up to 15 seconds, at a lower price than 1.1. See a real 15-second clip, the prompt behind it, and what each length costs in credits.

AI-assisted draft. Reviewed and edited by the Phosphene team before publication.

In brief

Gemini Omni 1.2 is now in Phosphene. Clips run from 4 to 15 seconds with dialogue and sound generated in the same pass, and 720p costs 41 credits per 5 seconds, down from 50 on Omni 1.1. Write the spoken line in quotes, give the scene its sounds, and save the end frame for clips of 10 seconds or less.

On this page
GEMINI OMNI 1.2 × PHOSPHENE lettering over a blurred night-market frame, with a blue NOW LIVE label and the line 15-second video with sound.
The closing card of our launch clip. The 15-second clip before it is one Gemini Omni 1.2 generation.

Google released Gemini Omni 1.2 Flash on November 5, and it is now in Phosphene. Two things changed from 1.1. Clips now run up to 15 seconds instead of 10, and each second costs less. Dialogue, music and sound effects still come out of the same generation as the picture.

Below is a real 15-second clip, the prompt behind it, and what each length costs.

One prompt, picture and sound together

A cook in a black apron at a neon-lit Bangkok night market grins at the camera while noodles toss in a flaming wok beside him.

Our launch clip is a single Gemini Omni 1.2 request at 720p, 16:9 and 15 seconds. The cook tosses noodles over a burst of flame, turns to the camera and says his line, plates the food and holds it out to the lens. The sizzle, the crowd and the scooter passing behind him are all in the soundtrack the model made. This is the prompt:

Handheld cinematic shot at a neon-lit Bangkok night market. A street food cook in a black apron tosses noodles high in a flaming wok, fire bursting up, then looks straight into the camera with a grin and says: "Fifteen seconds is just enough for one perfect plate." He slides the noodles onto a plate and hands it toward the lens. Loud sizzling, crowd chatter, a scooter passes behind him. Shallow depth of field, warm tungsten and pink neon light, steam in the air.

The same cook holds a plate of pad thai with shrimp and lime toward the camera, the night market glowing behind him.

Three habits made it work:

  • Put the spoken line in quotes and say who says it. Omni voices it and syncs it to the face.
  • Name the sounds you want as plain nouns: sizzling, crowd chatter, a passing scooter. They land in the mix.
  • Write the action in order. Fifteen seconds holds three or four beats, and Omni follows them in the order you list them.

Three ways to start

Gemini Omni 1.2 sits in the video model picker in three forms:

  • Video starts from text alone.
  • Animate starts from one photo as the first frame. Add a second photo as the last frame, and the model moves between them. With an end frame, clips run up to 10 seconds.
  • Reference builds the shot from up to 10 images and up to 3 short videos, for a product or a character you want to keep consistent.

All three output 16:9 or 9:16, at 720p or 1080p.

What it costs

Phosphene prices each clip from what the provider charges us. Gemini Omni 1.2 costs 18% less per second than 1.1, and the credit prices follow:

Length720p1080p
5 seconds41 credits62 credits
10 seconds82 credits124 credits
15 seconds123 credits186 credits

On Omni 1.1 a 5-second 720p clip was 50 credits. The composer shows the exact quote before you generate.

Where Omni 1.1 went

Omni 1.2 replaces 1.1 in Phosphene. Saved model picks and drafts that used 1.1 open on 1.2 automatically. Google moves its own 1.1 traffic to 1.2 on December 3.

Sources

Share this guide
X LinkedIn