# Google 推出 Gemini Omni，一款創新 AI 模型，旨在改變影片創作。

*genai · news · 2026-05-19 · Times Now*

## Key points

- Google 的 Gemini Omni 可從任意組合的文字、圖片、音訊和影片中創作高品質影片。
- Omni 提供對話式影片編輯，每個指令可持續並累積多輪操作。
- Omni 中改進的物理模擬提升真實感，包括更佳的重力、流體動力學和動能運動。
- Avatar 功能讓用戶使用自己的聲音和外貌生成個人化影片內容。
- Gemini Omni Flash 初期僅對 AI Plus 及每月 100 美元的 AI Ultra 訂閱用戶開放。

Google has introduced Gemini Omni, an innovative AI model aimed at transforming video creation by allowing users to produce high-quality videos. Google has announced Gemini Omni, a new AI model that takes video creation to a level that's either genuinely exciting for creators or quietly terrifying for anyone who values original content online. Probably both, if we're being honest. The US-based tech giant is positioning Omni as its most significant leap in generative media since Nano Banana raised the bar on AI image generation roughly a year ago. This time, video gets the same treatment and the scope of what's possible is considerably broader than previous tools. What Gemini Omni Actually Does Google describes Omni as the point "where Gemini's ability to reason meets the ability to create" which sounds like marketing copy until you look at what it actually does. The model accepts any combination of text, images, audio and video as input and produces high-quality video output from a single unified model rather than a chain of specialised systems. Google says it's "natively multimodal from the ground up," which in practical terms means more coherent edits, fewer visual artifacts, and cleaner results when working across different media types simultaneously. The headline interaction pattern is conversational video editing, each instruction builds on the last, and previous directions persist across turns so your video evolves in a coherent direction as you refine it. Practical use cases Google highlighted include changing environments within a clip, reimagining camera angles, refining sequences iteratively, and generating explainer-style content from short text prompts. The company also specifically emphasised improved physics simulation, gravity, fluid dynamics, kinetic energy which is exactly the kind of detail that separates AI video that looks artificial from footage that actually holds up. Google says users can create videos featuring their own voice and likeness through an Avatar feature, a digital version of yourself that generates videos that look and sound like you. Whether that's a powerful creative tool, a significant privacy concern, or simply the world's most efficient slop generator remains to be seen. Probably all three at once. Availability And Pricing Gemini Omni Flash is rolling out now inside the Gemini app for subscribers on AI Plus and AI Ultra tiers. Google's new $100-per-month AI Ultra plan includes priority access alongside bundled Omni Flash access. The model is also coming to Google Flow and YouTube Shorts. Get Latest News live on Times Now along with Breaking News and Top Headlines from Technology Science and around the world. Latest News Sparsh author Sparsh has been diving deep into the worlds of technology, science, and gaming for Times Network, bringing readers the latest updates and insights. Hi ... View More News Technology & Science End of Article

**Companies:** Google
**Countries:** United States

[Read the full story on Times Now](https://www.timesnownews.com/technology-science/gemini-omni-explained-googles-new-ai-model-that-can-create-video-from-any-input-article-154354400)

---

Canonical: https://newsio.io/zh-TW/n/a330e578-2c34-41e5-b29f-9b0fced997a1/google-gemini-omni-ai
Summarized by Newsio from Times Now. https://newsio.io/how-it-works