Gemini Omni 各版本与参数
| 参数 | Gemini Omni 1.1 Flash |
|---|---|
| 时长 | 3–10s |
| 画面比例 | 16:9 · 9:16 |
| 分辨率 | 360p · 720p · 1080p · 4K |
| 原生音频 | ✓ |
| 图生视频 | ✓ |
| 参考图生视频 | ✓ |
| 积分/秒 | 7.3 |
| 5 秒片段消耗 | 37 cr/5s |
什么是 Gemini Omni?
Gemini Omni 1.1 Flash is a Google video model that takes text, a still image, or a set of reference images and short video clips, and returns a video clip with its sound already in it. It was added to FlyAIgh in September 2026 alongside Google's older VEO 3.1 line, and it works differently from VEO in three practical ways.
First, length. Clips run any whole number of seconds from 3 to 10, in 16:9 or 9:16, instead of VEO's fixed 8 seconds. Second, references. Besides text-to-video, image-to-video and first-and-last-frame control, Gemini Omni has a reference mode that accepts up to 10 images plus up to 3 video clips of up to 3 seconds each, so a product that turns in your hand or a camera move can be shown rather than described. The reference clips do not add to the price. Third, sound. Every clip comes back with an audio track, and there is no switch: it is always on and already in the price. In our own test generations each file carried a stereo track.
Output resolution runs from 360p to 4K, and the 4K tier is a genuine 3840×2160 render rather than an upscale. Price follows resolution: 7.3 credits per second at 720p, so a 5-second clip is 37 credits and 10 seconds is 73. 360p costs about a quarter of that (20 credits for 10 seconds), 1080p about 1.5 times (113 for 10 seconds) and 4K three times (219 for 10 seconds). The cheap 360p tier is useful for testing a prompt before paying for the finished resolution.
What it does not do: there is no 1:1 or 3:4 framing, so square and near-square placements need another model, and no single clip runs past 10 seconds.
On FlyAIgh, Gemini Omni runs from the same credit wallet as every other model, with no Google Cloud project, API key or quota request, and the exact cost is shown before you generate.