Gemini Omni 1.2 Flash
Google's video model that makes and edits clips of up to 15 seconds with sound, and the successor to Veo in the Gemini app and the Gemini API.
Quality
Modality
video
Resolution
4K (drafts at 360p or 720p, upscaled to 1080p or 4K)
Access
closed
Fabian's Take
"I don't fully understand Google's product strategy and naming, but it looks like Omni replaces Veo. I've tried it, and it's quite decent. Right now, though, Seedance 2.5 seems to be the one dominating video."
Gemini Omni is Google’s newer video model, first shown at Google I/O in May 2026. Version 1.2 Flash came out on 5 November 2026. It makes clips of 3 to 15 seconds, up from 10 seconds in version 1.1, with sound generated in the same pass. Drafts come out at 360p or 720p, and you upscale the ones you like to 1080p or 4K. Every clip carries Google’s invisible SynthID watermark.
Omni replaces Veo
In the Gemini API, Google says so outright; elsewhere it’s less clear. In the Gemini app, videos now come from Omni. In the Gemini API, the Veo 3.1 preview models shut down on 22 October 2026, and Google named Omni Flash as their replacement. On Google’s Vertex AI platform, the finished Veo 3.1 versions stay available until at least May 2027, and in Flow, Google’s video tool, you can still choose between both. If you’re starting now, start with Omni.
Editing by conversation
The main difference from Veo is that Omni edits. You make a clip, then ask for a change, such as a different product on the table, other light or another camera angle, and it changes that and keeps the rest. That’s how you’d work with a person, and it saves generating a whole new clip because one detail was off. Google is holding back one editing feature for now: changing what people in a video say.
What it costs
$0.07 per second of 720p video through the Gemini API and Google Cloud, 30% less than version 1.1. Higher resolutions cost more. In Google’s apps it runs on your plan: Flow offers it to everyone, with direct scene creation and edits; in the Gemini app, 15-second videos are rolling out to Google AI Ultra subscribers first.
Where it stands
Quite decent, in my use. It’s the obvious choice if you already work in Gemini or Flow. If video quality is what matters most, Seedance 2.5 seems to be ahead right now.
The Verdict
Best for: Short clips with sound that you want to keep refining in conversation, inside Google's apps or through the Gemini API.
Pros
- Makes and edits clips of 3 to 15 seconds, up from 10 in version 1.1
- Conversational editing: change one thing in a clip and keep the rest
- Sound in the same pass, with clearer lip-sync and several speakers
- Takes text, images and video as input (audio references only for Google Cloud customers Google has approved)
- Cheaper than Omni 1.1: $0.07 per second at 720p
Cons
- Still a short-clip model: you can extend a clip to 30 seconds in steps, anything longer means stitching clips together
- Full 15-second clips in the Gemini app only for Google AI Ultra subscribers so far
- Google's naming is confusing: Veo, Omni and Flow overlap, and it isn't always clear which model you're using
Specs
- Pricing $0.07 per second at 720p (30% less than Omni 1.1 Flash) · other resolutions priced separately
- Cost Tier moderate
- ⚡Speed Tier fast
- License Proprietary