
What happened
The model update promises more natural frame dynamics and expanded control over generation.
Why it matters
The addition of native support for the vertical format eliminates a significant limitation for content creators focused on mobile devices, while improved motion realism could accelerate the adoption of AI video in media production.
Google DeepMind has announced the release of an update to its video model, designated version 3.1. According to the developers, the new version is capable of creating livelier and more dynamic clips that are perceived as natural and engaging.
A key functional addition is the support for generating video in a vertical format. This change expands the tool's scope of application, adapting it to the needs of modern mobile platforms and social networks.
Developers also emphasize improvements in frame consistency and the provision of users with finer control over the creative process of video content creation.
Facts
- Google DeepMind released an update to the Veo model under version 3.1.
- The new version supports video generation in a vertical format.
- Developers state that the update ensures higher consistency and control.
- Generated clips are described as lively, dynamic, and natural.
Context
Information is based exclusively on the meta-description of the official Google DeepMind blog post from January 13, 2026. Independent confirmations of technical specifications or examples of the model's operation are absent in the provided source package.
What remains unknown
- What are the specific technical differences in the architecture of Veo 3.1 compared to previous versions?
- Is the new model available to the general public or only to a limited circle of testers?
- How exactly is the improved control over generation implemented in practice?
AI analysis
The focus on vertical video indicates a strategic shift in the development of generative models toward integration with the short-video ecosystem (Shorts, TikTok, Reels). The mention of 'naturalness' and 'dynamics' suggests that previous versions may have suffered from static qualities or motion artifacts that are now being prioritized for correction.
Strategic AI conclusion
A likely consequence will be increased competition among video generation tools for market share in mobile content. The next observable signal should be the appearance of demo reels from the developers themselves or initial partners. Material uncertainty remains regarding the actual quality of generation and the tool's availability to end users.