Gemini Omni Flash: make and edit AI video from one prompt

Gemini Omni Flash is a video generation model from the Gemini family. In Flash Omni it is the default model of the workspace on the home page, and it covers three jobs that other tools usually split into separate products: turning a text prompt into a clip, turning still images into a clip, and editing a clip you already have.

Flash Omni is an independent workspace that runs the model for you. You do not need a Google account, an API key or any local setup. You open the page, describe the shot, and the workspace chooses the right mode from what you attached: a prompt alone runs text to video, a prompt plus images runs the reference mode, and a source video switches to editing.

Gemini is a trademark of Google. This site is not affiliated with or endorsed by Google; the name is used here only to say which model runs behind the workspace.

The model is built around short, self-contained clips, which is what most social, ad and storyboard work needs. Common first uses:

Every run exposes the same small set of controls, so there is not much to learn. You pick the aspect ratio (16:9 or 9:16), the resolution (720p, 1080p or 4K), the clip length (4, 6, 8 or 10 seconds) and optionally a seed. A fixed seed makes it easier to compare two prompt variations, because the only thing that changed is the wording.

For references, the image modes accept up to 7 PNG, JPG or WebP files of up to 10 MB each. The editing mode takes one MP4, MOV or WebM source clip of up to 100 MB plus up to 5 image references. Plan on roughly two minutes per clip, a little longer for edits; the exact time varies with load.

Flash Omni grants 300 credits when you create an account, and no credit card is required. A daily check-in adds 5 more credits on days you come back. That is enough for 3 short test clips at 720p or 1080p (90 credits each for a 4-second clip), or 2 clips at the default 6-second length.

Short prompts that name one subject, one action and one camera idea outperform long lists of adjectives. A reliable structure is: who or what is in the shot, what happens, how the camera behaves, and what the light and mood feel like. For example: a ceramic coffee cup on a wooden table, steam rising, slow push-in, soft morning window light.

Change one variable at a time. If you rewrite the subject, the camera and the lighting together you cannot tell which change helped. Keep the seed fixed while you tune wording, then release it when you want variety.

Draft cheap, finish once. Test composition with a 4-second 720p clip, and only render the 10-second or 4K version after the framing is right. The cost table in the pricing guide shows how much that saves.

Clips are short, between 4 and 10 seconds, so longer videos are built by generating shots one after another and cutting them together in your editor. Fine text, hands and complex physical interactions can still come out imperfect; if a result misses, adjust the prompt or attach a reference rather than repeating the same prompt.

Generations use credits, and credits from a failed generation are returned to your account automatically. Subscription credits refresh each billing cycle and expire if unused; one-time credit packs never expire. Current plan prices are always on the pricing page, never in this guide, so they cannot go stale.