# Gemini Omni 1.2 Flash Gemini Omni 1.2 Flash in Sloosh: Google video with sound, 3 to 15 seconds from 360p to 4K, with first and last frames. Last updated: 2026-11-06 **Key facts** - Gemini Omni 1.2 Flash is Google's newest Omni video model in Sloosh, beside Gemini Omni 1.1 Flash, and it always makes sound. - Gemini Omni 1.2 Flash makes 3 to 15 second clips in Sloosh at 360p, 720p, 1080p or 4K, in 16:9 or 9:16; a clip with start and end frames is up to 10 seconds. - Gemini Omni 1.2 Flash takes up to 6 reference images and 1 video of up to 10 seconds, or a video to edit of up to 15 seconds, in Sloosh. - Gemini Omni 1.2 Flash can extend a clip in 10-second steps up to 30 seconds in total, with the Video edit node's Extend task in Spacelab. Web page: https://sloosh.ai/docs/gemini-omni-1-2-flash Gemini Omni 1.2 Flash is Google's newest Omni video model. It makes longer clips than 1.1 (up to 15 seconds), edits videos of up to 15 seconds, and offers a choice of resolution (360p drafts up to 4K) and first-and-last-frame clips. It always makes sound. ## When to use it - Ads with people, sound and natural-language edits that need more than 10 seconds in one shot. - Changing a clip you already have, of up to 15 seconds. - A transition between two images you already have (up to 10 seconds). - Draft at 360p, then make the final at 1080p or 4K. ## How to prompt it - Write timed beats like `[0-5s]` and `[5-12s]`. - Describe what you want, not what you don't. - With first and last frames, describe only the change between them. - References work best for faces and likenesses; a reference video for style or camera alone doesn't work. [Prompting Gemini Omni](https://sloosh.ai/docs/prompting-gemini-omni.md) has the full guide, including captions, audio and tags. ## Example **Prompt** ```text A single continuous 15-second shot. Photoreal. A sunny kitchen counter, morning light through the window behind. The camera pushes in slowly on the blender from the first image as a hand pours a thick pink smoothie from its jug into a glass, then lifts the glass toward the window, ending close on the glass. Shallow depth of field, fresh and warm. Audio: the blender winding down, the pour, soft acoustic guitar. ``` ## Settings in Sloosh - Aspect ratio: 16:9 or 9:16. - Resolution: 360p, 720p, 1080p or 4K. - Length: 3, 5, 8, 10 or 15 seconds. - References: up to 6 images and 1 video of up to 10 seconds. - Video to edit: up to 15 seconds. - Start and end frames: both optional (an end frame needs a start frame); a clip with frames is up to 10 seconds. - Sound: always on. - Prompt: up to 50,000 characters. Open it in [Video](https://sloosh.ai/studio/video) and pick Gemini Omni 1.2 Flash in the model chip. The Generate button shows the cost first. ## Frequently asked questions ### How long can a Gemini Omni 1.2 Flash video be? A single Gemini Omni 1.2 Flash clip is 3 to 15 seconds in Sloosh, and a clip with start and end frames is up to 10 seconds. To go longer, extend it in 10-second steps up to 30 seconds with the Video edit node's Extend task in Spacelab, or join separate clips with Video Concat. ### How is Gemini Omni 1.2 Flash different from Gemini Omni 1.1 Flash? In Sloosh, Gemini Omni 1.2 Flash makes clips of up to 15 seconds and edits videos of up to 15 seconds; [Gemini Omni 1.1 Flash](https://sloosh.ai/docs/gemini-omni-1-1-flash.md) stops at 10 seconds for both. 1.1 extends a clip up to 40 seconds in total, 1.2 up to 30. Both offer 360p to 4K, 16:9 or 9:16, and always make sound. ### Can Gemini Omni 1.2 Flash use a start and end frame? Yes. Gemini Omni 1.2 Flash takes an optional start frame and an optional end frame, and an end frame needs a start frame. A clip with frames is up to 10 seconds. Describe only the transition between the two images. ### Does Gemini Omni 1.2 Flash use the audio from a reference video? No. Google says any audio in a Gemini Omni video reference is ignored, and that video references work best for likenesses such as a face. Write the sound you want on its own Audio line in the prompt. ## Sources - [Gemini API: Omni](https://ai.google.dev/gemini-api/docs/omni)