Google Releases Gemini Omni 1.2 Flash for Cheaper AI Video Generation

Google has launched Gemini Omni 1.2 Flash, a new artificial intelligence video generation model that can create 15-second videos, edit existing clips through conversations, and maintain greater consistency between scenes. The model is now generally available through the Gemini API and Gemini for business, featuring a 30 percent price reduction compared with the previous version. Google has reduced the cost of generating 720p videos to $0.07 per second, meaning a full 15-second clip costs approximately $1.05 through the API. The company is also bringing the model to Google Vids, Flow, Google Ads, and the Gemini app.

The release marks a significant expansion in Google media tools, offering developers and creators longer video lengths and more precise controls without increasing expenses. Businesses and individual creators can now draft content at lower resolutions before upscaling selected projects to higher formats. This approach aims to streamline digital production workflows across various enterprise and consumer platforms.

Specifications and Features of Gemini Omni 1.2 Flash

The new model supports longer video generation, improved reference accuracy, and editing through follow-up instructions. Users can generate videos ranging from 3 to 15 seconds in duration, with adjustments available in 1-second increments. Initial drafts can be rendered at 360p or 720p, while final outputs can be upscaled to 1080p or 4K. The system supports multi-turn conversational editing, motion references of up to 3 seconds, and image references for characters, products, and styles. Audio features include dialogue, multi-speaker voices, and lip-sync capabilities. Camera controls allow for flythroughs, zoom-outs, and orbits, while all outputs include an invisible SynthID watermark for identification.

Creators can combine character photographs, product images, style references, short motion clips, and audio cues together in a single workflow. For example, users can combine a model photograph with multiple clothing images to create a virtual try-on video showing different outfits from several angles. Google demonstrated another example in which the model repeatedly changed a video background while preserving the characters movements, clothing, and camera framing. It also showcased continuous camera movements, including a virtual drone flight through the Pantheon and a zoom-out from a shoreline to an aerial island view.

What Has Improved Over Gemini Omni 1.1 Flash?

Gemini Omni 1.2 Flash increases the maximum video duration from 10 to 15 seconds and introduces improvements in video consistency, speech generation, and editing. Users can change backgrounds, products, lighting, weather, or aspect ratios through follow-up prompts without generating an entirely new video. Google also states that the model better preserves faces, clothing, product packaging, text, and colors during movement and camera transitions. Compared to Omni 1.1 Flash, the new generation reduces visual drift and minimizes unwanted cuts during complex camera movements.

Dialogue generation and multi-speaker audio have also seen notable upgrades, allowing the system to match different voices to individual characters with natural timing. The integration of image and motion references helps maintain character continuity across multiple scenes in a 15-second clip. Developers using the API benefit from enhanced reliability when chaining multiple prompts together to refine specific visual elements.

Pricing Structure and Cost Reductions for Developers

Google has reduced the price of 720p video generation by 30 percent, moving from $0.10 per second down to $0.07 per second. The new pricing allows developers to produce longer videos at a lower cost per second. Generating a 3-second video now costs $0.21, compared to $0.30 previously. A 5-second clip drops from $0.50 to $0.35, while an 8-second video costs $0.56 instead of $0.80. A 10-second video is priced at $0.70 down from $1.00. Longer formats previously unsupported are now available, with 12-second videos costing $0.84 and 15-second clips priced at $1.05. These figures are calculated using stated 720p API generation rates and exclude separate upscaling or other applicable charges.

By allowing users to generate drafts at lower resolutions before upscaling, Google provides a more economical path for testing creative ideas. Instead of generating a completely new video whenever something needs changing, users can refine an existing clip through conversational editing. According to Google, these workflows help reduce production costs for marketing campaigns and independent media projects.

Availability Across Platforms and Enterprise Integration

Google is making Gemini Omni 1.2 Flash generally available through its API and enterprise platforms while expanding access across consumer and creative applications. The model is accessible via the Gemini API, Google AI Studio, Gemini for Business, and Google Cloud with enterprise support. In consumer and productivity software, users with a Google account or Google Workspace subscription can generate up to three free 15-second videos per month in Google Vids. The model is also available in Google Flow for video generation and conversational editing, while Gemini app access is rolling out to Google AI Ultra subscribers. Google Ads now supports video campaign creation using the technology.

Several external platforms have integrated or tested Gemini Omni within their workflows, including Adobe Firefly, Figma Weave, and Magnific, alongside organizations like Monks, Whering, Resemble AI, and STAN. Businesses can use the model through Google Cloud, including Provisioned Throughput options for applications requiring dedicated processing capacity. Google notes that while the model handles complex camera movements and character consistency well, it still requires ongoing improvements in fast sports action, rapid scene changes, and complex physics.