Compare Gemini Omni Flash quality tiers: Draft, Standard, and HD resolutions for text-to-video and image-to-video generation.
Pick the resolution and duration that fits your project — same Gemini Omni Flash model underneath.
Model selection guide
The models page is not just a price table. It explains how Omni groups video generation options by quality tier, duration, image support, and credit cost so you can choose the cheapest useful setting for each stage of production.
For most prompts, the cheapest useful workflow is to test direction at a lower tier before spending on higher-resolution output. Draft generations help you evaluate subject, motion, lighting, and camera language. Once the idea works, move to a standard or HD tier for the version you plan to publish or show to a client.
Text-only models are best for visual exploration when you do not have assets yet. Image-input models are better when the subject must remain recognizable, such as a product photo, character, interior, or brand visual. Choosing the right input type usually matters more than choosing the highest resolution immediately.
AI video quality often improves through iteration, not a single perfect prompt. A model that lets you test more variations may be more valuable during early ideation than a higher-cost setting. Omni exposes credit cost clearly so you can decide whether a clip is for exploration, approval, or final export.
Use lower tiers for prompt discovery, standard tiers for everyday social clips, and HD tiers for final-looking assets. Pair the tier with the use case: fast hook testing, product motion studies, landing-page visuals, or polished ad creative. This keeps generation choices tied to business outcomes rather than specs alone.
Start with Draft for fast iterations, switch to 720p or 1080p when you need higher quality, or use image-to-video when you already have a reference image ready.
Open generator