For teams making short video ads or product clips, the costs of AI video come from two places: the price per second and the number of times you have to regenerate a clip because one detail came out wrong. Google's Gemini Omni 1.2 Flash, now generally available in the Gemini API and Gemini for business, attacks both.
The headline changes are length and price. Clips can now run up to 15 seconds, up from 10 seconds in Omni 1.1, and duration can be set in one-second steps from 3 to 15. Google has cut the price by 30% from Omni 1.1, to $0.07 per second at 720p. The suggested workflow is to draft cheaply at 360p or 720p, fix details with follow-up edits and only then upscale the best version to 1080p or 4K.
Multi-turn editing is the feature that changes the bill. Instead of starting again, you can ask in conversation to swap a product or background, change the lighting or weather, or switch the aspect ratio, across a full 15-second clip, which makes producing variants for different audiences far cheaper. Google also claims better adherence to references: you can combine character, product or style images, short motion clips of up to three seconds and audio cues, and faces, garments, packaging text and brand colours are meant to stay consistent across camera angles. Dialogue, lip-sync and multi-speaker scenes are said to be more natural, and camera moves such as long drone-style flythroughs should have fewer unwanted jump cuts. Outputs carry an invisible SynthID watermark, and Provisioned Throughput is available for business customers.
Our Gemini Omni Flash review already ranked it as the cheapest way to edit a video by conversation, if not the best-looking model, and this release addresses the two complaints we had about 1.1, clip length and price. Our pricing notes put it at about $0.023 per second at 360p and $0.209 at 4K, which sits within the range of rivals such as Runway, at $0.02 to $0.21 per second by resolution.
Treat the quality claims with care. The consistency and lip-sync examples are Google's own showcase clips, chosen to look good. Google itself says work continues on fast sports action, rapid script changes and complex physics, which are the cases where video models still fail most visibly. Anyone using it for branded work should test it against their own product photography and character sheets before switching, and note that there is no free tier.