Compose a clip from a description alone, from up to seven reference pictures, or from one reference clip. Choose 4, 6, 8 or 10 seconds and 720p, 1080p or 4K; with a reference clip the length is decided for you.
The selector lists two Gemini Omni entries, and this page describes one of them: Gemini Omni Video, its own product rather than a setting of the other. It has no first-and-last frame mode and offers three quality tiers instead of four. Text mode needs only a description; in the image workspace you supply reference pictures, a reference clip, or both.
Pictures and the clip share one quota of seven units, and the clip counts as two, so seven pictures fit without a clip and five with one. The workspace uses up to the first ten seconds of a reference clip whose source is no longer than 30 seconds, and the output length is then decided automatically.
Choose 4, 6, 8 or 10 seconds only when the task carries no reference clip, and 720p, 1080p or 4K in every mode. The shape selector offers 16:9 or 9:16 and stays available with references. A description of up to 20000 characters is required, and an optional seed is available.
Combine pictures of a product, a setting and a character, then describe how they belong together. The clip is composed from what you supply, so exact framing and identity are not guaranteed and deserve a check.
Attach one reference clip, optionally with up to five pictures, and describe the action. The first ten seconds of the source are used and the output length is chosen automatically.
Without a reference clip you choose 4, 6, 8 or 10 seconds and 720p, 1080p or 4K. An optional seed lets you repeat a run with the same settings.
Supply pictures of the same product or character together with a location, describe the scene, and check that the subject still reads as the one your audience knows.
Use a clip you own as a reference, add up to five pictures, and describe the action. Output duration is chosen automatically and may differ from the reference segment.
Choose the resolution for your delivery needs. Comparing a lower-resolution result with 4K requires separate generations; keeping the description or seed does not guarantee identical motion or framing.
Use a button on this page. It opens the right workspace with this model already selected, in the language you are reading now.
Write what should happen in the shot or image. Concrete subjects, framing and action give the model more to work with than short labels.
Choose length or quality tier, shape and sound where offered, then read the credit quote for those exact settings before you generate.
Generation uses account credits. Cost depends on the model, quality and output size or duration. Reference images, video or audio may add charges in some modes. Check the workspace quote for your selected inputs and settings before generating; the pricing page lists credit packages.
See credit packagesThey are two separate products in the same selector. This entry has no first-and-last frame mode and offers 720p, 1080p and 4K, while the other adds a lower tier and a frame mode. Neither is presented here as the better one.
Seven per task. A reference clip takes two of those seven units, so the ceiling drops to five pictures when a clip is attached.
Yes: one clip plus up to five pictures, because the clip takes two of the seven units. Only one clip fits in a task, and no length can be set alongside it.
The workspace uses up to the first ten seconds of a source that is no longer than 30 seconds. Output duration is automatic and is not guaranteed to match that segment.
Yes, in every mode, up to 20000 characters. Describe what should happen, and in reference mode also say how the pictures belong together.
There is no sound on or off switch, and voice or character material cannot be uploaded for this entry here. Check the delivered file for a soundtrack and mute it in your editor when you need silence.
The figure follows the length and the tier you pick, and a task with a reference clip is quoted as one generation instead of by length. The workspace shows the amount for your settings before you confirm.
Other models on this platform that do a similar job, or that pair well with this one.
Create video from text, first and optional last frames, or reference images and a video. Choose 4, 6, 8 or 10 seconds without a reference video; otherwise output duration is automatic.
Read the guideGenerate from a description or animate a starting picture. Set any whole length from 2 to 15 seconds, choose 720p or 1080p, and pick from five shapes in text mode. This page covers the two generation entries.
Read the guideGenerate from a description or from a starting frame. Set 4, 6 or 8 seconds, choose 720p, 1080p or 4K, and pick a shape in both modes. This page covers the two Lite entries.
Read the guide