Describe one scene, choose its length, and create a short motion study.
Gemini Omni Flash 1.1
Gemini Omni 1.1 Flash genuine short test · 360p · 4s
A small red ceramic teapot on a cream table, subtle steam drifting upward, fixed camera, soft daylight, no lettering.
A still image suggests a moment; a short video shows how that moment unfolds. Explore a gentle camera move, a changing reflection, or a small environmental action with Gemini Omni 1.1 Flash on NanoPic. Use the teapot preview as a simple starting point, then replace the prompt with a scene of your own.
Use a compact shot to decide what should move, what should stay still, and where the viewer should look.
Describe a subject, its surroundings, one action, and one camera instruction. A quiet scene can be as useful as an elaborate reveal: a reflection crossing a bottle or a curtain moving beside a window makes it easy to judge timing. Keep competing actions out of the first draft so you can tell which instruction needs refinement.
Use image-to-video when a starting composition matters, or supply first and last frames to guide the intended transition. Choose frames that share a plausible subject, lighting setup, and viewpoint. Reference images can also communicate visual direction. Guidance is not a guarantee of exact reproduction, so inspect the subject through the entire clip.
Start at 360p to inspect the shot. Choose 720p or 1080p when detail matters, or 4K output for an upscaled delivery option. A larger file cannot repair a warped object or a confusing action. Resolve those issues in the prompt and references first, then judge a higher-resolution result on the screen where it will be used.
Set a direction, inspect the output, and refine with a specific next step.
Write down what viewers should notice at the opening and what should be different at the end. Select text-to-video for an open-ended concept; use frames when you already know the composition. Choose landscape or portrait before generation rather than hoping a later crop will preserve your subject.
The page starts with 360p, 4 seconds, and 16:9. Replace the example prompt with one focused instruction and review the displayed credit amount before submitting. Available lengths are 4, 6, 8, and 10 seconds. Increase duration only when the action needs more breathing room, not simply because a longer option exists.
Watch once for the overall idea and again for frame-level problems. Check edges, shadows, reflections, and whether the camera stops where you intended. Change either motion, framing, or lighting in the next attempt; changing all three makes comparison harder. Save the approved prompt and settings alongside your downloaded clip.
Adapt these suggested briefs; they are prompt ideas, not additional generated examples.
Try: A small glass vase on a wooden windowsill, a sheer curtain moving gently beside it, locked camera, overcast daylight, one continuous shot, no lettering. The curtain supplies the movement while the vase remains a stable reference. If the composition wanders, simplify the surroundings before adding more detail.
Try: A handmade paper landscape under warm desk light, the camera slowly approaches a tiny cardboard doorway, shallow depth of field, keep the paper texture and silhouette stable. This is a prompt idea, not a displayed result. Begin with one camera move instead of adding characters, weather, and changing scenery at once.
Plan around a real use, and review the finished clip before publication.
Test how light and atmosphere change a familiar still-life scene. A short motion study can explain a feeling to collaborators more clearly than a collection of adjectives.
Create one visual hook with room around the subject. Add captions in your editor after reviewing the video, rather than relying on generated lettering to carry important information.
Explore the movement between two related compositions. Treat the result as a planning reference and check whether the transition is believable before building a longer sequence around it.
For a second interpretation of the same short brief, explore Kling O3. For a broader set of supported reference workflows, visit Seedance 2.5. Keep subject, aspect ratio, and intended action consistent when comparing outputs, and judge the result you actually receive rather than a model name alone.
It is a Google video-generation model for translating written direction and visual inputs into moving scenes. This page focuses on short text-to-video, frame-guided, and image-reference workflows that you can control in a browser.
The generator offers 4, 6, 8, or 10 seconds and 360p, 720p, 1080p, or 4K output. Landscape 16:9 and portrait 9:16 help you plan for the destination. Start with a short draft, then select the format your project actually needs.
Treat it as 4K output using upscaling, not a promise that every detail was generated natively at that resolution. Review the delivered clip for fine texture and edge artifacts; output dimensions alone do not establish visual quality.
Yes, the image-to-video workflow accepts frame guidance. Use a clear starting image and, when needed, an ending frame that describes a plausible transition. Exact object shape and identity still need review, especially when the two frames differ substantially.
No. This generator is for new short clips from text and supported image inputs. It does not provide scene extension, uploaded-video editing, audio-reference input, or a multi-shot timeline. Plan a longer film as separate clips and assemble approved material in your editing tool.
Yes. The preview is a genuine 360p, four-second text-to-video test generated with this model. Its prompt is shown so you can see the simple motion brief. It demonstrates that example only, not a guarantee about every subject or higher-resolution setting.
Yes. Check the amount beside the generation control after choosing your settings. Welcome credits can contribute toward generation, but their availability does not mean unlimited free video or that every selected configuration is covered.
The generator on this page defaults to “Gemini Omni Flash 1.1”. The visible model selector is the source of truth for the model you will use; if you switch it, the selected model and its available controls change accordingly.
Start with the editable prompt above, keep the action simple, and review one short clip.
