Skip to main content
AI News & Updates

Gemini Omni 1.2 Flash Adds 15-Second Video: What Creators Should Validate First

Native duration rises from 10 to 15 seconds with multi-turn editing; verify pricing, the Google Vids allowance, and output consistency separately

ENHE AI5 min5 views

Google launched Gemini Omni 1.2 Flash on 2026-11-05 with native clips from three to fifteen seconds, multi-turn conversational editing, stronger reference following, and a listed API price of $0.07 per second for 720p video after a 30 percent reduction. A separate Google Vids announcement says Google Account and Workspace users can generate up to three fifteen-second videos per month at no cost. This guide turns those facts into one bounded creator test: prepare a simple reference pack, generate a single shot, inspect identity and product details at several points, make one controlled edit, record the real cost or allowance used, and keep a shorter fallback if the longer clip drifts.

More topic tags (7)
AI NewsAuto PublishingSEOGemini Omni 1.2 Flash15秒AI视频Google Vids免费额度多轮视频编辑
Gemini Omni 1.2 Flash Adds 15-Second Video: What Creators Should Validate First

Key takeaways

Google confirms native durations from three to fifteen seconds and multi-turn conversational editing.
The API post lists $0.07 per second for 720p, while Google Vids separately offers up to three fifteen-second videos per month at no cost.
Validate identity, product text, audio, and the final frame with one shot and one controlled edit before expanding.

By ENHE AI · 2026-11-09 · AI-assisted drafting and editorial checks.

Direct answer: start with one single-shot trial

Omni 1.2 Flash raises the native limit from ten to fifteen seconds and supports continued conversational edits on the same clip. Do not begin with a multi-character, multi-cut production. Choose one person, one product, or one explanatory action and prove that the beginning, middle, and end all preserve the brief.

This article is based on two Google announcements. It did not sign in to your account and does not treat vendor demos as your output. Interface, region, plan, and allowance must be checked in the actual account.

Confirmed duration, price, and entry points

Google says creators can choose durations from three to fifteen seconds in one-second increments and lists $0.07 per second for 720p video. A separate Google Vids post says Google Account or Workspace users can make up to three fifteen-second videos per month at no cost. API pricing and the Vids allowance are different scopes and must not be summarized as free access everywhere.

The post also describes reference images, short motion references, audio cues, and turn-by-turn changes to background, lighting, weather, or aspect ratio. Availability and actual credit use still need to be read in the selected product surface.

A team discusses a video workflow at a computer. Contextual stock photo, not an official demo or hands-on result. Image: Unsplash.
A team discusses a video workflow at a computer. Contextual stock photo, not an official demo or hands-on result. Image: Unsplash.

A reviewable input and procedure

Prepare one authorized main reference, one sentence describing the action, a fixed aspect ratio, and three details that must not change. Example: “Fifteen-second locked shot: the person lifts a blue mug and looks at camera; preserve the clothing pattern, mug text, and desk position; add no new people.” This is a task example, not a quality guarantee.

After generation, inspect frames near zero, five, ten seconds, and the end for identity, object count, text, hands, background, and audio. Change only one variable in the next turn, such as the background. If unrelated details drift, stop instead of stacking more edits.

Cost, failure handling, and rollback

Record the surface, duration, resolution, generation count, edit count, actual charge or allowance use, and human correction time. A free allowance only describes a specific surface; it does not make production cost zero. Check API and other product surfaces separately.

If the latter portion changes a person, clothing, product, or text, preserve the failed sample, fall back to eight to ten seconds or split the idea into two shots, then compare acceptance. Do not hide instability with more paid retries or publish an unreviewed clip.

Practical impact for everyday creators

Fifteen seconds can hold a complete product action, short explainer, or social ad, but it lengthens the consistency check. Multi-turn editing can avoid regenerating an entire clip, yet it may introduce unrelated changes. The useful saving is an accepted revision, not a lower count of button presses.

Before publishing, verify input rights, likeness permission, brand text, audio, platform disclosure, and the final export. This article does not promise ranking, conversion, or consistent model behavior across all assets.

FAQ

Can every Google user generate unlimited fifteen-second videos for free?

No. The Google Vids post says up to three per month. The API has per-second pricing, and other surfaces depend on account and product terms.

Does multi-turn editing guarantee consistent people and products?

No. Google describes stronger reference following, but your assets still require frame-by-frame checks and a shorter or split-shot fallback.

Related tools and reading

These pages support the same task, workflow, or decision boundary.

Conclusion

Treat fifteen seconds as a new limit to validate, not a promise of automatic stability. Complete one single-shot trial, record inputs, output, edit, cost, and rollback, then decide whether it belongs in a real publishing workflow.

Watch on the source website

What this means for everyday users

Record the surface, duration, resolution, generation count, edit count, actual charge or allowance use, and human correction time. A free allowance only describes a specific surface; it does not make production cost zero. Check API and other product surfaces separately.

Related reading

AI Video Generators vs Online Video Editors: How Beginners Should Choose

AI video generation tools such as Sora and Veo focus on creating new video from prompts or images. Online editors such as Canva and CapCut focus on timelines, captions, transitions and publishing workflows.

vLLM-Omni links real-time image generation and asynchronous video generation on SageMaker

AWS published a September 28 workflow for image and video generation with vLLM-Omni on Amazon SageMaker AI. Two endpoints are deployed from the same Deep Learning Container: FLUX.2-klein-4B runs as a real-time image endpoint, while Wan2.1-VACE-1.3B runs asynchronously for video. The generated still image and a motion prompt are submitted to the video endpoint, and the resulting MP4 is stored in Amazon S3, with an optional Streamlit interface for demonstration. The shared container simplifies framework packaging, but production teams still need separate capacity, queue, timeout, moderation, content-rights, retention, and idle-cost controls for the real-time and asynchronous stages, plus traceability from every final video back to its prompts and source asset.

Gemini Launches a Small-Business Hub for Brand Assets, Connected Apps, and Finance

Google has introduced a dedicated business landing page in the Gemini app for small-business tasks such as brand building, marketing, finance analysis, and customer insights. The announcement also lists connections to Google Business Profile, Ads, Workspace, and third-party apps including Canva, Shopify, Xero, and others. This article explains the global rollout wording and offers a practical two-part workflow: prepare a brand brief for a weekend promotion, then turn verified expense data into a reviewable chart. It emphasizes minimal app connections, source checking, and human approval before publication or financial use. The release can reduce context switching, but it does not mean every account receives the hub immediately or that generated claims, graphics, and numbers are ready to use without review.

From Chat Boxes to Personal AI Companions: AI Assistants Are Entering the Desktop Execution Era

AI assistants are moving from answering questions toward continuing real tasks. AI agents, MCP tool ecosystems, personal memory, and local workbenches are pushing this shift together. For users, the real value is not another chat box, but less repeated context setup and more continuity from thinking to doing.

ChatGPT Adds an Accounting Plugin: Keep Source, Variance, and Approval Tables for Month-End Work

OpenAI announced the ChatGPT Accounting plugin on November 18, 2026. It can help reconcile transactions, draft invoice entries, and prepare month-end reports from records and templates, while Enterprise workspaces still require administrator approval before installation. This article does not provide personalized accounting or financial advice. Instead, it supplies three control tables: a source table for files, periods, and versions; a variance table for every unmatched amount; and an approval table that names who may confirm, export, or roll back a change. Bank access, payments, credentials, and final posting remain human-reviewed steps. The plugin output is a reviewable draft, not a closed accounting period or evidence that the underlying records are complete.

GitHub Copilot Evals Enters Preview: Regression-Test Skills, Agents, and MCP Changes

GitHub launched Copilot Evals in public preview on November 17, 2026. It creates repeatable tests for skills, custom agents, and MCP tools and lets users compare runs before and after a customization change. This guide defines a minimum regression set: choose three realistic tasks, then specify the input, allowed tools, success criteria, and prohibited behavior for each. Preserve at least one baseline run and one post-change run, including the result, execution path, failed checks, and human review. Merge a customization only when the change preserves the original task, permission boundary, and evidence chain. A single passing answer is not enough because model output can vary, and an automated check cannot decide every domain-specific quality or safety question.