Gemini Omni 1.2 Flash

Google put a longer, cheaper AI video model into everyday tools on 5 November 2026 with Gemini Omni 1.2 Flash, generating clips up to 15 seconds long at 720p for $0.07 per second. Google states that rate is 30 percent lower than Gemini Omni 1.1, a change aimed at teams producing AI video in volume for marketing or product demos. The model handles native video generation, meaning it produces footage directly rather than assembling stock clips, along with conversational editing across several turns and improved reference following so output stays closer to supplied images and characters. In Workspace, Google's announcement says anyone with a Google account or Google Workspace subscription can generate up to three 15-second videos per month in Vids at no cost, under conditions the report notes are often understated.

Google DeepMind introduced Gemini Omni 1.2 Flash through a blog post on 5 November 2026, adding native 15-second video creation and conversational editing that lets creators refine clips turn by turn. Developers can reach the model through the Gemini API and Google AI Studio, with enterprise access via Google Cloud. Pricing is set at $0.07 per second for 720p generation, roughly a 30 percent reduction over Gemini Omni 1.1, and users can draft at 360p or 720p. The model supports continuous clip generation and editing up to 15 seconds long, and is available for creating and editing videos in Flow.

Google has released Gemini Omni 1.2 Flash, a generally available video-generation model that produces clips of 3 to 15 seconds in one-second increments and supports successive natural-language edits. Google states that 720p output costs $0.07 per second, roughly 30% below Omni 1.1, and the model is available through the Gemini API and Google Cloud. The launch materials describe output options from 360p up to 4K, with support for text-to-video, image-to-video, reference generation, frame interpolation, editing, extension, upscaling and synchronized audio workflows. Google lists a 131,072-token input limit and a 260,640-token output limit, while one independent 20-call test found most jobs finished in 40 to 90 seconds without being a standardized benchmark.

Google made Gemini Omni 1.2 Flash, its AI video model, generally available on 5 November 2026, cutting price by 30% to about $0.07 per second of 720p video through the Gemini API, down from $0.10 in June. Clips now run up to 15 seconds in one generation pass, up from 10 seconds. Users can edit a clip across several chat turns, swapping products, backgrounds, or lighting instead of regenerating from scratch. Google says faces, clothing, packaging text, and brand colors stay more consistent across angles, and DeepMind testers preferred 1.2 over 1.1 for visual consistency 62% of the time. The model rolls out in Flow, Vids, Ads, the Gemini app for AI Ultra subscribers, and Google Cloud.

Gemini Omni 1.2 Flash (google/gemini-omni-1.2-flash) is described as Google's native multimodal video generation and editing model and the successor to Gemini Omni 1.1 Flash. It produces clips up to 15 seconds long with native audio, supports 16:9 and 9:16 aspect ratios, and renders from 360p drafts up to 4K resolution, with a stated release date of Nov 5, 2026. Supported workflows include text-to-video, image-to-video, first and last frame interpolation, and subject and video references with improved reference management. The model can also edit or extend an uploaded video, and multi-turn editing and extension is handled by passing the previous job id, while video generation is only accessible through the Video Generation API.

Gemini Omni 1.2 Flash accepts text, image, and video inputs and returns video with synchronized sound, with generated and edited clips ranging from 3 to 15 seconds in one-second increments, compared with a 10-second ceiling in Omni 1.1 Flash. Clips created from a first and last frame are limited to 10 seconds, and Google says each output carries an invisible SynthID watermark. Hands-on testing showed that a color change to a single object altered only that element while preserving frame timing, whereas a stop-motion restyle changed textures but kept smooth motion and showed slight audio drift across edits. Extension limits differ by surface, with the Gemini API capping finished video at 30 seconds while Google Cloud documents up to 40 seconds total.

Google DeepMind published the official model card for Gemini Omni 1.2 Flash on 5 November 2026, describing it as a step toward models that can create and edit anything from any input, starting with video. It combines Gemini's intelligence with Google's generative media models, improving world understanding, multimodality, and editing. The card specifies that Gemini Omni 1.2 Flash accepts text strings, images, audio, and video as input and outputs high-resolution video with audio. The model is built on Gemini Omni Flash, with training data, hardware, and software details delegated to that base model card. Distribution covers the Gemini App, Gemini Enterprise App, and Google channels.