https://x.com/i/article/2105378350352998400
Creative prompting with Gemini Omni in Google Flow
To help get you started exploring Gemini Omni Flash in Google Flow, here is a comprehensive guide featuring prompt templates and detailed ideas for your next creative generations. Whether you are producing films, crafting marketing content, designing visual assets, or simply having fun experimenting with novel ideas, these tips will help take your outputs to the next level.
1. Provide high-level constraints
Instead of micro-managing every visual element, give Gemini Omni high-level creative constraints. The model excels at filling in artistic nuances like lighting consistency, set design, and physical nuances when instructed to pay attention to them.
Prompts to try:
- “Be extremely detailed in environments. Apply costume design principles to characters, specifying exact textures, props, and lighting.”
- “Consider micro-detail, facial expression, and timing to create a natural dialogue.”
- “Handheld fisheye camera shot of the character in the room.”
2. Use visual anchors
Establish your visuals to get precise color grading, action, and sequences with first and last frames. Uploading or generating a static frame defines composition, palette, and subject before the camera starts moving, enabling Gemini Omni to focus on the movement of the camera and subjects in the shots while ensuring the least amount of drift or hallucinations.
Prompts to try:
- “Smooth, continuous high-speed drone pull-back shot using this start and end frame. The camera rapidly retreats backwards and tilts down, ascending several hundred feet into the air. The mist parts to reveal the massive mountain range.”
- “Hyperlapse with a locked-off, static camera using this start and end frame. The transition is seamless, shifting the ambient lighting from cold, wet twilight to warm morning sunlight.”
- “Create an endless loop using this start and end frame.”
3. Leverage tagged ingredients
You don’t need to rely solely on text for your prompts, in fact, you have more creative control when you use additional ingredients. You can tag images for characters, settings, or style, upload video clips for audio or motion reference, and even use full storyboards. Simply name your assets as you add them to your workspace and then include them to your prompts directly.
Prompts to try:
- “Use [@ Tagged Image] as the product in a video ad campaign, where the product is being opened and tested by a woman set against a warm pink studio.”
- “Using this video as a motion and audio reference, change the subject to [@ Tagged Character], keep everything else the same.
- “Animate this [@ Tagged Image] storyboard image sequence, following key beats exactly in order from top to bottom, left to right.”
4. Refine pacing mid-scene
Sometimes it is helpful to get the desired action and then edit for the desired pacing. Once you have your generation, update the motion using simple descriptive language. While Gemini Omni will understand ‘fast’ or ‘slow,’ getting more granular yields better results.
Prompts to try:
- “4x the moving speed of the interaction between the two subjects.”
- “As the dog reaches the hoop, time ramps down into extreme slow motion.”
- “Fast-walking character; at sec 3, a sky-falling flower nears her hand, then time slows to 1% speed as it slips through her fingers. Single shot, no cuts.”
5. Transfer and mix styles
Blend existing aesthetics onto new subjects or environments. Think of Omni as your stylist for each shot, give it your visual references and describe what pieces you want to use from it.
Prompts to try:
- “Apply the pose and motion from this video to the character in this image reference.”
- “Apply the visual look of this photo to the generation, replacing all materials with the appearance of the video reference.”
- “Match the color grading of this image to my scene.”
6. Render animated text
Omni generates native high-fidelity kinetic typography. You can control font styles, placement, alignment, and movement pacing. Perfect for captioning your content or creating tutorial content.
Prompts to try:
- “Add hand-drawn marker-style text annotations and arrows for each ingredient in this cooking video.”
- “Word-by-word on screen: ‘Did you know this model can do typography?’ Each word appears with a unique animated style, synced to a sizzle reel rhythm.”
- “A speech bubble appears next to the woman reading, ‘If I fall asleep now, how many hours of sleep will I get?”
7. Enforce continuous scene rules
By default, Gemini Omni attempts to cut between multiple angles to create narrative depth. If you want a continuous "oner" shot, you must explicitly state it in your prompt.
Prompts to try:
- “The camera follows the characters down service stairs, through a bustling kitchen, and into the crowded nightclub.”
- “Shot as a continuous, flat side-scroller tracking shot along a field. The camera moves laterally along a rail as a single protagonist does show jumps on horseback.”
- "Without any scene cuts, capture a single continuous, handheld shot of the character walking up to the door.”
8. Introduce transitions and cuts
While Gemini Omni introduces cuts and camera angle changes natively, directing the cuts can help create more engaging transitions and narrative depth. Try to describe clear sequence steps for cinematic multi-shot transitions like this example below.
Prompts to try:
- “Fast transit lights streaking across a dark city highway transforming into bioluminescent waves crashing on a night beach.”
- “A match-cut sequence of a spinning coffee cup, a massive Ferris wheel, and a close-up of an analog watch face.
- “A close-up shot of a woman blowing up a balloon, followed by a cut to her rowing on a lake, ending with a wide shot of the lake filled with balloons as she rows
9. Precision conversational editing
When tweaking existing video clips, keep edits hyper-focused. Elaborate descriptions can accidentally trigger unintended environmental changes. Use the phrase "Keep everything else the same" at the end of your editing prompt.
Prompts to try:
- “Make the phone in her hand invisible. Keep everything else the same.”
- “Change the sign text to read ‘Omni Flash’. Keep everything else the same.”
- “Add a dramatic rim light on the subject while maintaining character features.”
10. Direct with timecodes
Gemini Omni outputs native audio synced to video elements. You can direct exact timestamps and beat-syncing using simple natural language or timecode brackets.
Prompts to try:
- “Include a high-energy synth beat synced to the visual transitions.”
- A locked camera shot of a beautiful suburban sidewalk.
[0-3s] A cat runs from left of frame and off frame right
[3-6s] A man chases after the cat from left of frame and off frame right [6-10s] A group of dogs are chasing the man and the cat from left of frame and off frame right
- Character ice skating in a pink ice ring like the image attached: At 2 seconds, the character enters from the camera left. At 5 seconds, it starts snowing and the character looks up.
And don’t forget to add negative keywords to your prompts to define what you don’t want to be included in your generations. Prevent unwanted background clutter, dialogue, or extra audio effects by specifying simple structural negative constraints. Here are some to try:
- No dialogue
- No scene cuts
- No extra sound effects
- No plastic textures
- No camera movement
Ready to get started? Try out these tips to bring your ideas to life.