Gemini Omni

Positioning Overview

The text-to-video route of Gemini Omni Flash is designed for developers who need to produce finished videos directly from plain text prompts. You only need to provide a single description covering the scene, action, camera movement, dialogue, and sound to generate a video with a native audio track. You can specify a duration of 4 to 10 seconds, a resolution from 720p to 4k, and landscape or portrait aspect ratios—making it ideal for rapid production and batch test shoots of short ads, vertical social media assets, and narrative storyboards.

Core Highlights

• A single prompt covers visuals and sound—the prompt can describe the scene, action, camera movement, dialogue, and sound all at once, producing a finished video with synchronized visuals and audio in one call.
• 4k resolution and dual landscape/portrait ratios—output supports three resolution tiers (720p, 1080p, 4k) and two aspect ratios (16:9, 9:16), so the same prompt can cover both landscape placements and vertical short videos.
• seed parameter ensures reproducibility—passing a fixed seed lets you rerun the same prompt, making it easy to compare prompt changes or establish a stable baseline for batch generation.

Capabilities

Text-only video production: a single prompt (up to 20,000 characters) is enough to generate a video, with no asset uploads required—ideal for validating creative ideas from scratch.

Adjustable duration and resolution: output duration can be selected among 4, 6, 8, and 10 seconds, and resolution ranges from 720p to 4k, so you can choose a tier based on placement and quality requirements.

Landscape/portrait aspect ratio switching: aspect_ratio supports 16:9 and 9:16, so the same creative idea can directly produce two versions suited to long-form video platforms and vertical feeds.

Reproducible generation control: seed accepts integers from 0 to 2147483647; once fixed, regenerating the same prompt is repeatable, which helps with version comparison and regression validation.

Use Cases

• Ad and marketing teams validating multiple copy versions—generate several candidate short videos from the same prompt, quickly compare shots and pacing at 4 to 10 seconds, and invest in polish only after settling on a direction.
• Social media operations producing vertical short-video assets—choose 9:16 and 720p or 1080p output for daily update assets on short-video platforms.
• Storyboard previews and script visualization—screenwriters or directors can put scene descriptions and camera movements into the prompt and get preview clips with dialogue and sound for proposal review.
• Presentation scenarios requiring high-resolution finished videos—for quality-sensitive presentations or large-screen use, 4k output can be selected to reduce subsequent upscaling.

Limitations and Selection Advice

Only text input is accepted; reference images or existing videos cannot be uploaded for conditional control. If you need to maintain subject consistency or transform existing assets, choose the image-to-video or reference-to-video route in the same family. Output duration is limited to 4 to 10 seconds, and aspect ratios are only 16:9 and 9:16, so continuous long shots or special frame formats outside this range are not suitable for this route. This model is currently in a coming-soon state; availability and actual generation results are subject to the official launch. For long narrative projects, it is recommended to validate with short clips first before expanding.