Table of Contents
Gemini Omni 1.2 Flash is Google's updated video generation model, released into general availability on November 5, 2026 with up to 15-second clips, stronger reference adherence, and a 30% lower official 720p API rate. Because Google's official release includes developer and ecosystem features that are not identical to what AI Studio currently supports, this guide separates the official 1.2 upgrade from what you can actually use on this site today.
The November 5, 2026 Release at a Glance
Google launched Gemini Omni 1.2 Flash into general availability on November 5, 2026, replacing the August 27, 2026 release of Gemini Omni 1.1 Flash as the current generation in the Flash video lineup. Rather than remaining behind a narrow waitlist, the 1.2 update arrived for production workflows across Google's official developer channels and cloud tools.
The update targets three practical limits from 1.1 Flash: maximum clip duration, per-second generation cost at standard 720p resolution, and how reliably the model follows visual references across a shot. Where Gemini Omni 1.1 Flash topped out at 10 seconds officially, the 1.2 specification supports continuous clips from 3 seconds up to 15 seconds in 1-second increments.
At the same time, keep one important distinction in mind when reading official release notes: Google AI Studio (Google's developer workbench) is not the same service as AI Studio at ai-studio.video. When Google announces new API parameters, resolution tiers, or conversational editing tools inside its own ecosystem, those capabilities do not automatically transfer one-for-one into third-party web interfaces. Below, we cover both the official model changes and the exact specifications available on this site.
Official Changes in Gemini Omni 1.2 Flash That Matter
Gemini Omni 1.2 Flash introduces several concrete upgrades at the model and official API level compared with the August 2026 1.1 Flash release. If you follow Google's official announcements, these are the core specifications behind the November 5, 2026 update.
15-Second Clips in 1-Second Steps
The most visible structural change in Gemini Omni 1.2 Flash is flexible clip duration. Officially, the model generates video from 3 seconds to 15 seconds in 1-second steps, up from the 10-second ceiling on 1.1 Flash. Being able to select an exact runtime in single-second increments—whether a 4-second B-roll insert or a 15-second continuous shot—reduces the need to pad prompts or trim extra frames when timing an edit.
30% Lower Official 720p Pricing
Google also reduced the official 720p generation price by 30 percent compared with Gemini Omni 1.1 Flash. In Google's official API pricing documentation, standard 720p output on Gemini Omni 1.2 Flash is $0.07 per second, down from about $0.10 per second on 1.1 Flash.
For developers tracking raw API economics, that reduction makes longer 10-to-15-second generations roughly comparable in cost to shorter clips under the previous pricing tier. However, these dollar-per-second figures apply strictly to Google's direct API billing—not to credit usage on ai-studio.video/pricing.
Stronger Reference Following and Consistency
According to Google's release announcement, prompt and reference adherence received a substantial upgrade in 1.2 Flash. When supplied with reference inputs, the model is designed to maintain steadier subject identity, wardrobe details, lighting direction, and scene geometry across the duration of a clip. In Google's official ecosystem, the broader reference pipeline also accepts additional media inputs, though—as detailed in the next section—only still image references are supported on this website.
Multi-Turn Conversational Editing and Official Ecosystem Features
Within Google's own tools and API environment, Gemini Omni 1.2 Flash supports multi-turn conversational editing, allowing users in those environments to refine a generated clip through follow-up instructions rather than starting every variation from scratch. Google's official release scope also covers 360p draft generation for lower-cost iteration, 1080p and 4K upscaling, scene extension, and video or audio reference uploads.
Every one of those capabilities belongs to Google's official platform toolset. None of those five features—360p drafts, 1080p/4K upscaling, conversational multi-turn editing, scene extension, or video/audio reference uploads—are available on ai-studio.video right now.
Early Arena Benchmarks (November 6, 2026)
On November 6, 2026, one day after the general availability launch, preliminary leaderboard results from public Arena blind comparisons placed Gemini Omni 1.2 Flash first in text-to-video with a score of 1541 (about 27 points above Gemini Omni 1.1 Flash), first in image-to-video, and third in video editing. Because these November 6, 2026 rankings were captured immediately after launch, they are preliminary figures and may change as more community votes accumulate.
Using Gemini Omni 1.2 Flash on AI Studio Right Now
On AI Studio's Gemini Omni page, Gemini Omni 1.2 Flash is now the default model, while Gemini Omni 1.1 Flash remains available on its own page for creators who still want it. Knowing the exact modes, durations, and credit rates in the web interface will help you plan your runs accurately.
Default Model Access, Modes, and Aspect Ratios
When you open the main Gemini Omni workspace, your session defaults to Gemini Omni 1.2 Flash with no external API key required. The site interface supports three generation modes and three aspect ratio settings:
- Text-to-video: Generate clips directly from written prompts.
- Image-to-video: Guide motion using one or two still images. Note that two-image image-to-video is capped at 10 seconds on this site, even though 1.2 Flash supports up to 15 seconds in other modes.
- Reference-to-video (up to 10 images): Supply up to 10 reference images to anchor characters, objects, or visual style across your generated clip.
- Aspect ratios: Choose 16:9, 9:16, or Auto depending on whether you are creating widescreen, vertical, or source-matched video.
Duration, Resolution, and Credits on This Site
Instead of charging Google's $0.07/s API rate, ai-studio.video uses site credits tied to your plan on the AI Studio pricing page. On this site, the exact duration, resolution, and credit rules are:
- Gemini Omni 1.2 Flash: Runs from 3 to 15 seconds at 720p only (with a default of 4 seconds, and a 10-second cap when using two images in image-to-video). It costs 7 credits per second at 720p.
- Gemini Omni 1.1 Flash: Runs from 4 to 10 seconds at 720p and costs 9 credits per second at 720p.
- Live cost display: The Generate button shows the exact credit cost before you submit a job, so you can verify your total as you adjust clip length.
What AI Studio Does Not Offer Yet
Even though Google's official 1.2 Flash documentation highlights them, you cannot do the following on ai-studio.video:
- No 360p draft mode: Output on the site is 720p only; 360p test renders are not offered.
- No 1080p or 4K upscaling: Higher-resolution upscaling tiers are not available here.
- No conversational multi-turn editing: Each generation is a standalone job; you cannot edit an existing clip through follow-up chat turns.
- No scene extension: You cannot extend a completed clip beyond its initial generation run.
- No video or audio reference uploads: While reference-to-video supports up to 10 still images, uploading video clips or audio files as references is not supported.
Gemini Omni 1.2 Flash vs. 1.1 Flash: Official vs. On AI Studio
Comparing Gemini Omni 1.2 Flash against Gemini Omni 1.1 Flash is easiest when you view the official model specifications side by side with what is exposed on ai-studio.video.
| Feature / Specification | Gemini Omni 1.1 Flash (Official) | Gemini Omni 1.2 Flash (Official) | On AI Studio (ai-studio.video) |
|---|---|---|---|
| Release Date | August 27, 2026 | November 5, 2026 (Current) | 1.2 Flash is default on /gemini-omni; 1.1 Flash still listed on /gemini-omni-1-1-flash |
| Clip Duration | Up to 10 seconds | 3 to 15 seconds (in 1-second steps) | 1.2 is 3–15 seconds at 720p (default 4s); 1.1 is 4–10 seconds at 720p; two-image image-to-video on 1.2 is capped at 10 seconds |
| Pricing / Credits | ~$0.10 per second (official 720p API) | $0.07 per second (official 720p API, 30% lower) | Site credits only: 7 credits/s at 720p on 1.2 Flash vs. 9 credits/s at 720p on 1.1 Flash (see AI Studio pricing) |
| Reference Inputs | Baseline 1.1 reference support | Stronger reference adherence + multimodal inputs | Up to 10 images in reference-to-video; no video or audio reference uploads |
| Resolution & Upscaling | Standard output tiers | Adds official 360p drafts and 1080p / 4K upscaling | 720p only (16:9, 9:16, or Auto); no 360p drafts and no 1080p/4K upscaling |
| Editing & Extension | Single-pass focus | Multi-turn conversational editing and scene extension | Single-run generation only; no conversational editing or scene extension |
Keeping Gemini Omni 1.1 Flash's August 27, 2026 history in view shows how quickly the Flash line moved: in just over ten weeks, the model expanded from a 10-second ceiling to 15-second shots while lowering both official API pricing and per-second credit costs on this site.
Should You Use 1.2 Flash or 1.1 Flash on AI Studio?
For most new projects on this site, starting with Gemini Omni 1.2 Flash is the practical choice, both because it is cheaper per second in site credits (7 credits/s vs. 9 credits/s on 1.1) and because it unlocks 3-to-15-second durations.
When to Start with Gemini Omni 1.2 Flash
If you are starting a new storyboard, building clips from up to 10 reference images, or generating shots longer than 10 seconds, stay on the default Gemini Omni 1.2 Flash page. In Google's announcement, 1.2 Flash is specifically built for stronger reference consistency and steadier subject tracking across longer clips, and on ai-studio.video it also uses fewer credits per second at 720p than 1.1 Flash.
If you are comparing how Gemini Omni fits alongside other video models hosted on the platform, you can browse the main AI video generator hub or read our guide to Sora alternatives.
When Gemini Omni 1.1 Flash Still Matters
Because Gemini Omni 1.1 Flash remains listed on its own page, you do not have to switch an active project mid-production. Keep 1.1 Flash in mind if:
- You are matching shots from an August–October 2026 project: Model updates can shift lighting contrast and pacing. If you created a sequence on 1.1 Flash and need one pickup shot to match existing renders, running that shot on 1.1 Flash keeps visual behavior consistent.
- You have existing prompts tuned for 1.1: Teams with established 4-to-10-second prompt templates built around the August 2026 model can continue using 1.1 until they are ready to transition.
Need Step-by-Step Prompting Help?
This article covers what changed in the November 5, 2026 release and what is supported on the site. If you want a step-by-step walkthrough on writing prompts, setting up image-to-video or reference-to-video inputs, and choosing settings, read our tutorial: How to Use Gemini Omni.
When you are ready to try the updated model with your own prompts or reference images, head to the Gemini Omni generator page.
Sources
All release dates, model capabilities, preliminary rankings, and pricing figures in this article were checked on November 10, 2026 against primary official documentation and platform specifications:
- Introducing Gemini Omni 1.2 Flash (Google Official Blog, November 5, 2026) — Primary release announcement covering general availability, 3–15 second clip durations in 1-second steps, stronger reference following, and multi-turn conversational editing in Google's tools.
- Gemini Omni 1.2 Flash Generally Available (Google Cloud Blog, November 2026) — Cloud and enterprise availability details, including official resolution tiers and ecosystem workflows.
- Google Gemini API Pricing Documentation — Official API pricing verifying the $0.07/second 720p rate for Gemini Omni 1.2 Flash compared with approximately $0.10/second for Gemini Omni 1.1 Flash.
- Build with Gemini Omni 1.1 Flash (Google Official Blog, August 27, 2026) — Historical release notes and 10-second baseline specification for Gemini Omni 1.1 Flash.
- AI Studio Platform Pages (Checked November 10, 2026): Site-specific support, 720p resolution limits, 7 vs. 9 credits/second pricing, reference-to-video (up to 10 images), and the 10-second two-image cap on Gemini Omni, Gemini Omni 1.1 Flash, and AI Studio Pricing.
Frequently asked questions
When did Gemini Omni 1.2 Flash launch?
Google released Gemini Omni 1.2 Flash into general availability on November 5, 2026, following the earlier Gemini Omni 1.1 Flash release from August 27, 2026. The 1.2 update expanded official clip duration from 10 seconds up to 15 seconds in 1-second increments and lowered the official 720p generation rate by 30 percent.
Is Gemini Omni 1.2 Flash available on AI Studio right now?
Yes. Gemini Omni 1.2 Flash is the default model on the Gemini Omni page at ai-studio.video. However, not every feature announced by Google is available here. Two-image image-to-video runs up to 10 seconds on this site, and official features like 360p drafts, 1080p or 4K upscaling, scene extension, and multi-turn editing are not offered.
How many credits does Gemini Omni 1.2 Flash cost on AI Studio?
On ai-studio.video, Gemini Omni 1.2 Flash runs at 720p for 3 to 15 seconds (with a 4-second default) and costs 7 credits per second. By comparison, Gemini Omni 1.1 Flash runs for 4 to 10 seconds at 720p and costs 9 credits per second. The Generate button always displays the exact credit total before you run a job.
Can I still use Gemini Omni 1.1 Flash on AI Studio?
Yes. While Gemini Omni 1.2 Flash is the default version on the main Gemini Omni generator page, Gemini Omni 1.1 Flash remains listed on its dedicated page at ai-studio.video. You can still select 1.1 Flash if you want to keep an existing workflow or stay consistent with prompts written for the August 2026 model.
Does AI Studio charge Google's $0.07 per second API price?
No. The $0.07 per second figure is Google's official API rate for 720p output on Gemini Omni 1.2 Flash. AI Studio at ai-studio.video is an independent web platform that uses subscription plans and generation credits rather than per-second API billing. You can view current credit allocations on the AI Studio pricing page.
What is the difference between Google AI Studio and ai-studio.video?
Google AI Studio is Google's own developer environment for testing Gemini models and API keys. AI Studio at ai-studio.video is a separate consumer web platform that provides access to multiple video generation models in one browser interface using a credit-based system, with its own mode, duration, and resolution boundaries.
Turn your idea into a video
Explore the video generator and start creating.
Generate a video


