- 3 to 15 seconds
Clip length
- 360p, 720p, 1080p, 4K
Resolution
- 16:9 or 9:16
Aspect ratios
- Text, first and last frame, up to 6 references
Inputs
- Native sound and speech, always on
Audio
About Gemini Omni 1.2 Flash
Gemini Omni 1.2 Flash is Google's newest Gemini Omni video model, generally available on Vertex AI since November 5, 2026, and Masonry runs it on Vertex directly. It generates video with synchronized native audio, including dialogue, from a text prompt, from a first frame with an optional last frame, or from up to six reference images of characters, products, or a style. Clips run any whole-second length from 3 to 15 seconds, up from 10 seconds in Omni Flash 1.1, in a 16:9 or 9:16 frame. Google says 1.2 brings more natural dialogue, clearer lip-sync, and better handling of several speakers in one scene.
For business teams, the practical change is range at a lower cost per second. One model now covers a 15-second spoken product moment, a vertical clip that keeps an approved packshot or presenter on screen, and a first-to-last-frame reveal. Draft at 360p or 720p, approve the direction, then render the keeper at 1080p or 4K, which Google produces by upsampling the same generation. Every clip carries Google's SynthID watermark. Omni 1.2 Flash replaces Gemini Omni Flash 1.1 in Masonry, and Google prices it about 30% below 1.1 per second at 720p.
Why teams choose Gemini Omni 1.2 Flash
Choose Gemini Omni 1.2 Flash for short, sound-on clips where speech, a real product, or a fixed opening and closing frame matter, at a lower price per second than its predecessor. It replaces Omni Flash 1.1 with longer clips, reference images, and better lip-sync. MiniMax H3 Max remains Masonry's everyday default for 5 to 15 second clips and lip-syncs 11 languages; Omni 1.2 Flash is the pick for a language H3 does not speak and when you need 4K output. For one unbroken take longer than 15 seconds, or to edit footage you already have, use Kling 4.0, which runs up to 30 seconds and edits uploaded clips. WAN 3.0 is worth comparing when native 1080p motion quality is the priority.
What Gemini Omni 1.2 Flash can do
The capabilities that set Gemini Omni 1.2 Flash apart and earn its place in a brief
Clips From 3 to 15 Seconds
Pick any whole-second length from 3 to 15, with 8 as Masonry's default. That is half again the 10-second ceiling of Omni Flash 1.1, enough for a complete spoken line or a short product story in one take.
Speech With Clearer Lip-Sync
Sound is generated with the picture, dialogue included. Google says 1.2 improves natural delivery, lip-sync, and scenes with more than one speaker, so write the exact line and who says it in the prompt.
Up to 6 Reference Images
Give the model up to six images of the people, products, or style to keep, with at most three images per character and two characters per clip. Say in the prompt which image is which and what each subject does.
First and Last Frames
Start from an approved frame, and optionally end on a second one; the model moves from the first to the last in order. A clip with both frames runs 3 to 10 seconds.
Draft at 360p, Deliver at 4K
Generate at 360p or 720p, with 720p as the default, or request 1080p or 4K. Google produces the higher two by upsampling, so they are larger outputs of the same generation rather than added native detail.
Physics and Camera in Plain Language
Direct pans, push-ins, orbits, and tracking shots by describing them, and the model grounds motion in Gemini's understanding of the real world, which suits pours, splashes, fabric, and hands in use.
Where teams reach for Gemini Omni 1.2 Flash
- Spoken product momentsA presenter or customer says one approved line on camera, with the product in shot, in a single 8 to 15 second clip.
- Ads that keep the real productPass approved packshots as references so the bottle, box, or device on screen stays the one you sell, then check every frame against the SKU.
- Vertical social clips with soundGenerate 9:16 clips for Reels, TikTok, and Shorts with ambience and speech already in the file.
- Reveals between two approved framesOpen on a closed box and land on the finished packshot, or move from a sketch to the final look, with the first and last frames fixed.
- Localized dialogue draftsDraft the same line in another language for a market test, then approve the script and pronunciation before anything ships.
- Cheap drafts before the final renderExplore pacing and camera at 360p, then render the chosen take at 1080p or 4K.
Frequently asked questions
What teams need to know about creating with Gemini Omni 1.2 Flash in Masonry
What is Gemini Omni 1.2 Flash?
Gemini Omni 1.2 Flash is Google's newest Gemini Omni video model, generally available on the Gemini API and Vertex AI since November 5, 2026. It generates video with synchronized native audio, including speech, from text, frames, or reference images. Masonry runs it on Vertex AI.
How is Gemini Omni 1.2 Flash different from Omni Flash 1.1?
Clips run up to 15 seconds instead of 10, you can pass up to six reference images, and Google says dialogue, lip-sync, and multi-speaker scenes are better. Google also prices it about 30% below 1.1 per second at 720p. Masonry retired Omni Flash 1.1 when 1.2 launched, and its page now redirects here.
What happened to Gemini Omni Flash 1.1 and the Omni Flash preview in Masonry?
Masonry retired the preview in November 2026 and Omni Flash 1.1 on November 5, 2026, when Omni 1.2 Flash replaced it. Both pages now redirect here, and videos you already made with them stay in your workspace.
How long can a Gemini Omni 1.2 Flash video be?
Any whole-second length from 3 to 15 seconds, with 8 as the default in Masonry. A clip with both a first and a last frame runs 3 to 10 seconds.
What resolutions does Gemini Omni 1.2 Flash support?
360p, 720p, 1080p, and 4K, with 720p as the default. Google produces 1080p and 4K by upsampling the generation.
Which aspect ratios does Gemini Omni 1.2 Flash support?
16:9, the default, and 9:16. There is no square option.
Does Gemini Omni 1.2 Flash generate audio and speech?
Yes. Sound is generated with every clip, including dialogue, effects, and ambience, and it is always on. Write the spoken line and who says it in the prompt.
Can I use reference images with Gemini Omni 1.2 Flash?
Yes. Pass up to six reference images of characters, products, or a style, with up to three images per character for up to two characters. References replace the first frame, so use one or the other.
Does Gemini Omni 1.2 Flash support first and last frames?
Yes. Image-to-video starts from a first frame you supply and can end on an optional last frame, and the model moves between the two in order. With both frames the clip is 3 to 10 seconds.
Can Gemini Omni 1.2 Flash edit or extend a video in Masonry?
Google's model can edit and extend clips, but Masonry does not offer those modes for it yet. To edit a video you already have in Masonry, use Kling 4.0.
Are Gemini Omni 1.2 Flash videos watermarked?
Yes. Google adds an imperceptible SynthID watermark to every video, and Vertex AI lists Content Credentials (C2PA) support for the model.
How much does Gemini Omni 1.2 Flash cost in Masonry?
Masonry bills in credits by resolution and clip length, and the editor shows the exact cost before you generate. Per second, it costs less than Omni Flash 1.1 did at every resolution.
Can I run Gemini Omni 1.2 Flash from the command line?
Yes. Install the Masonry CLI with npx @masonryai/cli, sign in with masonry login, and run masonry models params gemini-omni-flash-1.2 to see the inputs it accepts before you generate.
How does my team use Gemini Omni 1.2 Flash in Masonry?
Open a Masonry canvas, pick Gemini Omni 1.2 Flash from the model selector, and describe the video you need: a product shot, an ad creative, a social post. Masonry generates it, then you refine, edit, and combine Gemini Omni 1.2 Flash with other models in one workspace.
Is Gemini Omni 1.2 Flash free to try?
New Masonry accounts get 50 free credits to try GPT Image 2.5 Flare. Gemini Omni 1.2 Flash is included in every paid plan, and the editor shows the exact credit cost before you generate.
How do I write good prompts for Gemini Omni 1.2 Flash?
Write one scene like a short shooting script: the subject, the action, the camera move, the setting, and what is heard, with any spoken line in quotes next to the person who says it. With reference images, name each one's role in the prompt, for example "the woman from the first image holds the serum bottle from the second". With a first and last frame, describe the change that connects them rather than either frame. Set the length in the duration control, since it overrides any length written in the prompt, and say "no dialogue" when you plan to add a voiceover later. See the prompt gallery on this page for real Gemini Omni 1.2 Flash prompts you can copy and adapt.
Who makes Gemini Omni 1.2 Flash?
Gemini Omni 1.2 Flash is built by Google. Inside Masonry it runs alongside 50+ image and video models, so your team can pick the right one for each brief without switching tools.
Can I see examples made with Gemini Omni 1.2 Flash?
Yes, the prompt gallery on this page shows real videos teams have generated with Gemini Omni 1.2 Flash in Masonry, each paired with the exact prompt you can copy and adapt for your own brand.
Start creating with Gemini Omni 1.2 Flash
Generate, edit, and compare across 50+ models in one workspace.
Guides for Gemini Omni 1.2 Flash
Prompt walkthroughs and examples from the Masonry blog
Explore more AI models
Compare Gemini Omni 1.2 Flash with other models teams run in Masonry