Input
Text or image
Start from a prompt, or upload one image to guide the opening frame.
Your video will land here
Write a prompt, then generate
Gemini Omni 1.1 Flash
Gemini Omni 1.1 Flash turns a text prompt or one start image into a short 720p video with native audio. It is designed for fast multimodal generation, with the final clip length decided automatically by the model.
Input
Text or image
Start from a prompt, or upload one image to guide the opening frame.
Output
720p
This page uses the 720p tier for a balanced production cost.
Length
Auto · 3–10s
The provider decides clip length; there is no exact duration parameter.
Formats
16:9 · 9:16
Choose landscape or vertical output before starting generation.
What to know
The page keeps the basic workflow visible and moves provider-specific behavior into clear, separate explanations instead of a long block of prose.
Begin from a written idea, or upload one image when the opening subject, product, wardrobe, or composition should anchor the generated shot.
There is no manual duration control in this integration. Gemini Omni Flash decides the final clip length within its short-form generation range.
Choose 16:9 for wide compositions or 9:16 for phone-first creative before generation so the shot is composed for the target frame.
Capabilities
Use it for rapid iteration when text, image context, motion, and audio all belong in the same creative brief.
Turn a concise creative brief into a short video without preparing a source image first.
Use one start image to anchor the subject or opening composition.
Describe dialogue, ambience, music, or effects in the same prompt as the visual direction.
Generate 9:16 creative drafts for Shorts, Reels, and TikTok-style placements.
Test product shots, camera ideas, and visual hooks before committing to a longer production workflow.
Use the fixed-cost page to explore several directions without managing an exact clip duration.
How it works
The workflow stays simple even though the model itself supports more advanced multimodal controls.
Describe the subject, action, framing, camera movement, lighting, mood, and any audio direction.
Upload one image when the opening composition or subject identity matters to the result.
Select 16:9 or 9:16. The model decides the final clip length automatically.
Prompt ideas
Specific camera and sound direction usually produces a clearer creative brief than a vague one-line prompt.
Cinematic
“A cyclist crosses a rain-soaked city bridge at blue hour, cinematic tracking shot, reflections on wet asphalt, soft traffic ambience and distant thunder.”
Product
“A matte-black smartwatch floats above a glass pedestal, slow orbiting camera, clean studio reflections, restrained electronic sound design.”
Portrait
“Animate the person in the reference image turning toward camera and smiling naturally, preserve identity and clothing, subtle handheld motion, quiet room ambience.”
Vertical ad
“A creator opens a package in a bright apartment, quick social-video pacing, natural reactions, crisp close-ups, vertical framing, light room sound.”
Prompting
A concise shot brief is usually more useful than a long list of disconnected visual adjectives.
Tell the model what the main subject does during the shot. A single coherent action is easier to stage than several unrelated events competing for the same short clip.
Use concrete camera language such as locked close-up, slow tracking shot, overhead move, or handheld follow to shape how the action is presented.
Add location, time of day, lighting, weather, or production style only when those details change the look and mood of the scene.
Describe dialogue, room tone, music, ambience, or sound effects in the same prompt when audio is part of the intended result.
A start image is most useful when it establishes subject identity, product appearance, wardrobe, or the opening frame. Let the prompt describe movement and progression.
Choose 16:9 for landscape placements or 9:16 for vertical social content before generating so the composition is built around the target frame.
Input guide
The simplified workspace leaves advanced provider controls out of the first screen. Use the input image, prompt, and format choice deliberately.
Use a single scene with a clear subject and action. Short generated clips are easier to direct when the prompt has one visual center instead of several competing events.
Choose an image that establishes the subject, product, wardrobe, or opening composition you want. The prompt should explain what changes after that frame.
Describe the action in the order it should happen, then add camera and audio direction. This is clearer than mixing unrelated style words and actions together.
Choose landscape for wide compositions and vertical for phone-first content before generating so the subject placement works with the intended publishing surface.
Because the provider determines the final clip length, write a scene that can resolve naturally within a short 3–10 second window rather than depending on exact timing.
When a result is nearly right, adjust the most important missing instruction—camera, movement, lighting, or sound—so each new generation tests a clear hypothesis.
Use cases
The current page is optimized for quick, short-form generation rather than long editing sessions.
Prototype a visual hook, product moment, or opening shot before producing a full campaign asset.
Create quick vertical or landscape clips from a text idea or creator portrait.
Turn a still frame into a moving shot to test pacing and camera direction.
FAQ