Pick a mode
Choose text, image, reference, edit, or extend.
Omni 1.2 Flash
Create with Omni 1.2 Flash, Google’s omni video model. Generate from text, animate photos, follow references, or edit and extend your own clips in one place.
Model capabilities
Everything from one omni model
Start from text or media, or edit and extend a clip you already have.
Text to Video
Describe a scene in plain language and Omni 1.2 Flash turns it into a 3–15-second video with natural motion and detail in 360p or 720p.
Image to Video
With Omni 1.2 Flash, you can animate one or two photos into a scene while keeping faces and products recognizable.
Reference to Video
Combine 1–10 images and an optional short clip to guide Omni 1.2 Flash, helping characters, styles, and products stay consistent across shots.
Video Edit
To edit with Omni 1.2 Flash, upload one original MP4 (up to 15s, 50 MB) and describe the changes to outfits, backgrounds, or objects. You can add up to 10 reference images, and the output keeps the original duration.
Video Extend
Use Omni 1.2 Flash to add 3–10 seconds to an existing clip. The original can be up to 30 seconds, and you can add up to 10 optional reference images.
Pay-per-second credits
Billing is credits-based and priced per second of video. The exact estimate appears on the Generate button before you submit, and failed generations are refunded automatically.
Specifications
Omni 1.2 Flash specs
Modes, durations, resolutions, references, and credits — matching what NovaVideo supports today.
| Model | Omni 1.2 Flash |
|---|---|
| Creation modes | 5 — text-to-video, image-to-video, reference-to-video, video edit, video extend |
| Output length | 3–15 seconds (dual-image image-to-video and extend append capped at 10 seconds) |
| Resolutions | 360p and 720p |
| Aspect ratios | 16:9, 9:16, auto |
| References & @ tags | Up to 10 images for reference/edit/extend (2 for image-to-video). @image and @video in reference, edit, and extend. Original video: MP4 only, max 50 MB; edit ≤15s input, extend ≤30s, reference ≤10s. |
| Pricing | Credits based on resolution, output seconds, input video seconds, and image count — estimate shown on Generate before you submit |
Official demos
Omni 1.2 Flash in action
Four Google announcement demos show reference following, detail consistency through motion, text-guided edits, and continuous camera moves.
Follow people and products from references
In Google’s try-on demo, Omni 1.2 Flash uses clothing photos as references to show an outfit from multiple angles.
Keep details consistent through motion
See how Omni 1.2 Flash turns a static design into a motion graphic while keeping the lettering, materials, and lighting consistent.
Refine a clip with text instructions
Ask to change the background, lighting, or an object, and say what to keep. Google shows several background variations of the same clip. Here, you can submit a result as the input for a new edit.
Guide a continuous camera move
Describe a push-in, orbit, or fly-through for Omni 1.2 Flash. Google’s Pantheon demo follows one continuous path from the square into the building.
Demos by Google. Results vary with the source material and prompt. Read the Google announcement
How it works
How Omni 1.2 Flash works
Go from a prompt or source clip to a downloadable video in four steps.
Add your materials
Add images or an original clip as needed.
Describe the result
Describe what to create or change.
Generate and download
Generate your clip and download it when ready.
Use cases
What creators make with Omni 1.2 Flash
Explore ideas for product videos, character stories, social clips, and edits.
Product marketing
Use Omni 1.2 Flash to animate product photos into lifestyle scenes or keep the same product recognizable across a campaign.
Product launch teasersE-commerce ad creativesSeasonal campaign visuals
Character-driven stories
Give Omni 1.2 Flash character reference images to reuse the same character across shots.
Character intro clipsStory continuationsSeries intros
Social content
Use Omni 1.2 Flash to turn a photo into a short vertical video or restyle footage for feeds and stories.
Photo-to-video postsTrend remixesBefore-and-after reveals
Footage editing and extension
Edit your own clips with text instructions, or extend a short take into a longer sequence without reshooting.
Outfit and weather swapsScene extensionsBackground changes
Explore more
Discover Other AI Video Generators
Compare other video models available on NovaVideo.
Before you start
Omni 1.2 Flash FAQ
Find answers about uploads, references, video length, and credits.
Omni 1.2 Flash is a video model from Google DeepMind. On this page, you can create short videos from text, images, or reference materials, edit existing footage, and extend clips. Choose from five modes to match your materials and what you want to create.
Yes. Upload one original MP4 (up to 15 seconds, 50 MB) and describe what you want to change and keep. You can optionally add up to 10 reference images for clothing, products, or scene guidance. Editing follows the original clip’s duration.
Yes. Video editing and extension each accept one original video and up to 10 optional reference images. Reference-to-video requires 1–10 images and can include one optional reference video up to 10 seconds.
Upload your media and type @ in the prompt to select an image or video. This works in reference-to-video, editing, and extension. Explain each asset’s role, such as using a reference image to replace a product. Results can differ from the reference.
Yes. Upload an original MP4 up to 30 seconds long, choose 3–10 seconds of additional footage, and optionally add up to 10 reference images. Describe the next action and camera movement.
This page offers 360p and 720p. Text-to-video, reference-to-video, and single-image generation support 3–15 seconds; image-to-video with two images supports 3–10 seconds. Video editing accepts an original up to 15 seconds and follows its duration. Video extension adds 3–10 seconds. 1080p and 4K are not offered here.
Credit usage depends on the mode, output duration and resolution, and input materials, including video duration and the number of reference images. Check the estimated credits on the generate button before submitting.
No separate Gemini subscription or Google sign-in is required. Use an account on this site with enough credits to submit your task.
Novavideo is an independent video and image creation platform using third-party AI models. Learn more