HYPE OR FACT? · CHECKED 5 NOVEMBER 2026
Has Google launched Gemini Omni 1.2 Flash for native video with audio on the Gemini API?
OUR VERDICT
FactOur original verdict: “Yes, as a documented product launch. Google's developer blog (5 Nov 2026) and AI Studio docs name the model Gemini Omni 1.2 Flash (`gemini-omni-1.2-flash`): native video with audio, durations 3–15 seconds, multi-turn edit/extend via the Interactions API, up to six reference images, and a stated ~30% price cut vs Omni 1.1 (~$0.07 per second of 720p output on the paid tier per blog and pricing docs). Docs say it is generally available on the paid Gemini API / AI Studio path; Google Vids offers up to three 15-second videos a month at no cost; Gemini app access is rolling out to Google AI Ultra. Caveats that do not overturn the launch: editing/extending uploaded user video is unavailable in the UK, EEA and Switzerland (model-generated edits remain); API docs say uploading audio inputs/references is unsupported even though the promo video shows voice refs; blog and API docs disagree on Provisioned Throughput (blog says available; Omni docs say not supported) — we flag both, we do not guess which wins. We have not benchmarked quality or latency hands-on.”
THE CLAIM
“Introducing Gemini Omni 1.2 Flash, now available in AI Studio and the Gemini API: fast and 30% lower cost, longer clips, edit in conversation, more natural speech, more controllable shots.”
What’s actually true
- Official name and model id: Gemini Omni 1.2 Flash / gemini-omni-1.2-flash (blog, model page, DeepMind model card published 5 Nov 2026).
- Docs: native multimodal video+audio up to 15s; duration 3–15s in 1s steps (default 10s); Interactions API multi-turn edit/extend; up to 6 reference images.
- Pricing docs: paid-tier GA; ~$0.07/s effective at 720p (5,792 tokens/s at $12/1M video tokens) plus input tokens; 1.1 was ~$0.10/s — matching the blog's 30% cut and $0.07/s figure.
- 1.1 auto-routes to 1.2 on 3 Dec 2026 and bills at 1.2 rates (docs).
- Google Vids: up to three 15-second videos per month at no cost for anyone with a Google account or Workspace (Vids blog).
- Regional limits documented: editing/extending uploaded videos not available in EEA, Switzerland, UK; model-generated edits supported. Audio reference uploads unsupported in current API.
- Provisioned Throughput: developer blog says available; aistudio.google.com/docs/omni Limitations says not supported — conflict flagged, not resolved.
Sources
Cost
Paid Gemini API / AI Studio: about $0.07 per second of 720p video output plus input tokens (Google pricing page). Google Vids: up to three 15-second clips a month free. Gemini app: rolling out to Google AI Ultra (blog).
USE IT IF
Developers and creators who need documented native video+audio generation and conversational editing on Google's stack.
SKIP IT IF
Anyone who needs verified audio-reference uploads via the current API (docs say unsupported), UK/EEA/Swiss users who must edit their own uploaded clips, or anyone treating the 25s promo video as proof of average quality or latency.
How we knowChecked on 5 November 2026. Verdicts: Fact = the claim holds up; Partly fact = real, but with caveats or overstated parts; Hype = the headline claim isn't supported. Products change quickly, so check the maker's site before you buy. Our methods & sources ↗