Gemini Omni 1.2 Flash prompt guide
Gemini Omni 1.2 Flash makes clips of up to 15 seconds with speech and sound from one prompt. Write the scene, then what each person says, then what is heard around them.
What Gemini Omni 1.2 Flash takes
- Takes a prompt and a first frame.
- Clips of 3 to 15 seconds.
- Aspect ratios: 16:9, 9:16.
- Settings: 360p; 720p; 1080p; 4K.
- From 2 credits per second.
The order to write a prompt in
- The scene in one sentence: who is there, where, and what happens.
- What each person says, in quotes, and how they say it.
- The sounds around them and any music.
- The camera and the look.
Tips
- A clip is 15 seconds at most: about two or three short lines of speech. Longer speech gets rushed or cut.
- Put each line in quotes and say who speaks: a barista says cheerfully, "Your usual?"
- With a photo as the first frame, describe what happens and what is said; the model takes the look from the photo.
- Pick 9:16 in the settings for Reels, Shorts and TikTok, 16:9 for YouTube.
- Draft at 360p or 720p: 1080p and 4K are Google’s upscales of the same clip and cost more a second.
Example prompts
A young chef in a bright home kitchen holds up a jar of chilli oil and smiles at the camera. She says: "Three ingredients, five minutes, and it goes on everything." Oil sizzles in a pan behind her. Medium close-up, warm morning light.
Close-up of a phone on a café table lighting up with a notification. A narrator says calmly: "Your bank, finally on your side." Soft café chatter and a light piano melody. Slow push-in, shallow depth of field.
Two friends on a park bench in autumn. One asks, "Did you really quit?" The other laughs and says, "Best decision this year." Leaves rustle, distant kids playing. Static medium shot, golden-hour light.