Next version said to extend clips to 15 seconds
Google is developing "Gemini Omni 1.5," the next version of its video generation AI "Gemini Omni," @lyraxana, an X account that reports rumors about new models, posted on Oct. 18 (Oct. 19 Japan time). The length of video that can be made in a single generation will be extended from the current 10 seconds to 15 seconds, according to the post. Google has not announced it, and other specifications such as image quality, pricing and release timing are unknown.
Other articles also mention the name "Omni 1.5," but none is tied to a confirmed release schedule. In September, the operator of video editing app CapCut wrote that it could not confirm a release date, model documentation or pricing for Omni 1.5.
Current 1.1 Flash tops out at 10 seconds
Gemini Omni is a Google model that creates videos with audio from text, photos and videos, and can also edit them through conversation. It was announced at the "Google I/O" developer conference in May 2026 and replaced the previous "Veo 3.1" in the Gemini app. The current "Gemini Omni 1.1 Flash," released on Aug. 27, supports output of up to 4K through the developer API.
A single generation produces 3 to 10 seconds of video, and a feature that adds continuations in 10-second increments can extend it to a total of 40 seconds. At 15 seconds, longer scenes could be made without using extensions. The length of a single clip has become one of the main metrics for comparing video generation AI.
| Model | Length per generation (seconds) |
|---|---|
| Gemini Omni 1.1 Flash (current) | 10 |
| Gemini Omni 1.5 (leak) | 15 |
Companies update video models in quick succession
In video generation, other companies made announcements around the same time. Germany's Black Forest Labs announced "FLUX 3 Max" on Oct. 19 (Oct. 20 Japan time), saying it can make a 10-second video in 12 seconds. China's Kuaishou began accepting API access applications for its video generation model "Kling 4.0" on Oct. 21.
