Google’s Gemini Omni 1.2 Flash Makes 15-Second AI Video Generally Available

Google DeepMind made Gemini Omni 1.2 Flash generally available on November 5, 2026, extending native AI video generation and conversational editing to clips as long as 15 seconds and cutting the listed 720p price 30% to $0.07 per second. The model is rolling out on the Gemini API, Google Cloud, Gemini for business, Google Vids and related Google apps, according to Google’s product blog and a companion Google Cloud post. Those figures are company-stated product terms, not independent lab measurements. For related model coverage, see The Boston Report’s AI Models hub and prior reporting on Gemini 4 Carbon.

What changed from Omni 1.1

Google says Omni 1.2 Flash raises the native generation and editing ceiling from 10 seconds to 15 seconds and lets developers specify duration in one-second steps from 3 to 15 seconds. The company also describes stronger adherence to image, short motion-clip and audio references, multi-turn conversational edits that avoid regenerating an entire clip from scratch, and improved speech and lip-sync. Outputs carry an imperceptible SynthID watermark, Google says. The Verge’s November 5 report summarized the same length and editing claims from Google’s announcement.

By the numbers: price and length

At 720p, Google lists $0.07 per second for Omni 1.2 Flash and says that is about 30% below Omni 1.1. In TBR arithmetic, a 15-second 720p clip at the new rate costs $1.05, versus $1.50 at a prior $0.10-per-second equivalent implied by a 30% cut to $0.07. A 10-second clip falls from about $1.00 to $0.70 under the same assumption. Google also describes drafting at 360p or 720p and upscaling selected outputs to 1080p or 4K; resolution-specific prices beyond the 720p figure were not independently verified here beyond Google’s own posts. The chart below compares the company-stated maxima and the 720p list price expressed in cents per second.

Where it is available — and what is still limited

Google Cloud’s post says Omni 1.2 Flash is generally available for production use with Provisioned Throughput options. Google Vids offers up to three 15-second generations per month at no cost for users with a Google account or Workspace subscription, according to a separate Workspace blog post dated November 5. Google notes that audio references remain allowlisted and that complex physics, fast sports action and rapid script changes are still areas of ongoing work. Those caveats matter for buyers treating the GA label as a finished creative tool rather than a faster draft engine.

Why the release matters for model watchers

Omni sits at the intersection of Gemini’s multimodal stack and generative video. Google positions 1.2 Flash for product videos, virtual try-ons and ad variants rather than as a long-form film system. That product framing differs from cybersecurity-focused Gemini 4 Carbon, which TBR covered when Google began limited rollout to security firms. Readers tracking model releases can compare both lines on the Artificial Intelligence category feed and the Latest AI News archive.

Pricing math readers should redo for their own resolutions

Google’s public 720p figure anchors the chart in this article. Enterprises generating 1080p or 4K, or buying Provisioned Throughput, will see different unit economics. The Workspace free tier of three 15-second Vids generations per month is a consumer on-ramp, not a production budget. Agencies should model failure retries and multi-turn edits: conversational editing can reduce full regenerations, which is why Google emphasizes that workflow for cost control.

Sources

• Google: Introducing Gemini Omni 1.2 Flash (Nov. 5, 2026)
• Google Cloud: Gemini Omni 1.2 Flash generally available
• The Verge: Gemini Omni 1.2 Flash 15-second clips
• Google Workspace: Omni 1.2 Flash in Google Vids

Additional sources

• Google DeepMind: Gemini Omni 1.2 Flash model card