Google DeepMind's Gemini Omni 1.2 Flash: 15-second video and a 30% price cut
Quick answer
Google DeepMind announced on Nov 5 that Gemini Omni 1.2 Flash is generally available across the Gemini API and Gemini for business. The update raises native video length to 15 seconds (from 10 on Omni 1.1), adds multi-turn conversational editing, improves reference adherence, and cuts 720p pricing 30% to $0.07 per second (Google DeepMind blog, Nov 5).
What’s new
- Length: generate and edit clips natively up to 15 seconds; duration can be set in one-second steps from 3 to 15.
- Editing: refine the same clip turn by turn (swap product/background, change lighting or aspect ratio) without regenerating from scratch.
- References: combine character, product, or style images, short motion clips (up to three seconds), and audio cues with stronger consistency across camera angles.
- Workflow cost: draft at 360p or 720p, fix details with follow-ups, then upscale keepers to 1080p or 4K.
- Speech / camera: Google says dialogue lip-sync and continuous cinematic camera moves (FPV flythroughs, orbits) improved versus 1.1.
Why it matters
- Fifteen seconds is enough for a short product explainer or social ad in one generation, not a stitch of 10-second takes.
- Multi-turn edits cut the usual “regenerate the whole clip for one change” tax that burns credits.
- Outputs carry an imperceptible SynthID watermark; Provisioned Throughput is available on Google Cloud.
What we don’t know yet
- How well fast sports, rapid script changes, and complex physics hold up — Google flags those as still improving.
- Exact consumer availability timing across every Google app surface (Flow, Vids, Gemini app Ultra, Ads) beyond “rolling out.”
FAQ
- Is this live for developers today? Yes. Google says it is generally available on the Gemini API and Google Cloud.
- Do I have to start over to change one detail? No. Multi-turn conversational editing is meant for that.
- Will people know it is AI video? Google says outputs include an imperceptible SynthID watermark.