How do you get 4K output from Google Gemini Omni Flash?

Google Gemini Omni Flash outputs 720p by default; to get 4K, select the 4K resolution tier at generation time — it costs the equivalent of a full generation in credits. A 1080p upscale sits in between at no extra credit cost, making 720p the cost-free baseline rather than a capability ceiling.

Generate your clip as normal first — Omni Flash defaults to 720p because that is the zero-extra-cost tier, not because the model tops out there. From that baseline you have two upgrade paths: an upscale to 1080p at no additional credit cost, or the native 4K tier, which is priced at the equivalent of a full generation. In practice, requesting 4K roughly doubles what you spend on that clip.

Factor clip length into the 4K decision before you commit credits. Omni Flash generates at 4, 6, 8, or 10 seconds, and because the 4K surcharge scales with the generation itself, a 10-second 4K clip costs meaningfully more than a 4-second one. Lock your take at 720p first — check motion, framing, and timing on the cheap tier — then re-run or upgrade only the selects to 4K. Across 30+ test outputs in one benchmarking pass, this select-then-upscale sequence is what kept credit spend controlled.

Native 4K is worth paying for when the clip is delivery-bound, because no other current AI video model offers it natively — Veo 3.1, Kling, and Seedance 2.0 all require an external upscale pass to reach 4K, while Omni renders it directly. That said, resolution and texture quality are separate things: benchmarking found Omni Flash's visual textures are not yet at prime-time cinema standard, so 4K output gives you delivery-ready pixel dimensions, not automatically cinema-grade imagery. For social and web delivery in your film's aspect ratio, the free 1080p upscale usually covers you; reserve the 4K tier for big-screen or archival masters.

One verification note: these credit tiers come from hands-on benchmarking, and Google's pricing can shift — confirm the current per-resolution credit costs in the Gemini API documentation before budgeting a longer production.

These steps cover the resolution path — how far up the tiers you go depends on where the clip is being delivered.

Watch some of these to see what works for you:

Google's Omni model is one of the few models, if not the only model out there that offers native 4K. The moment we have a 4K model that has very very very strong VFX capabilities, we will finally have an AI model that is ready for big screen primetime.

— invideo's creative team