Google DeepMind announces Veo 3.1 update

Google DeepMind has released Veo 3.1, the latest version of its AI image creation model Veo. This update focuses on creating more vivid and natural video clips, and in particular, supports new vertical video creation functions.

According to Google DeepMind, Veo 3.1 has strengthened three key elements: 'consistency', 'creativity', and 'control'. The company explains that the generated video clips provide more dynamic, natural, and immersive results.

Why is this important?

Support for vertical video creation is a notable change in the field of AI video creation. As short-form platforms such as TikTok, Instagram Reels, and YouTube Shorts become the mainstream of global content consumption, demand for vertical videos is explosively increasing. Considering that most existing AI video creation models were optimized for horizontal (16:9) videos, Veo 3.1's vertical support is expected to provide practical value to creators and marketers.

In addition, strengthening 'consistency' suggests that problems such as inter-frame inconsistency, character deformation, and visual flickering, which were chronic problems in AI video generation, have been improved. This is an element that raises the possibility of commercial use of AI-generated videos to the next level.

Differences from previous versions

Google DeepMind did not disclose specific benchmark figures or technical specifications, but emphasized 'natural and engaging' results and promoted overall improvement in visual quality.

AI video creation competitive landscape

There is currently fierce competition in the field of AI video creation. OpenAI's Sora, Runway's Gen-3, Pika Labs' Pika 2.0, and China's Kuaishou's Kling are competing for the market.

Google DeepMind's Veo has been continuously updated since it was first released in 2024. In particular, when combined with Google's extensive cloud infrastructure, it is expected to provide strong competitiveness in the enterprise market.

Future outlook [AI analysis]

Veo 3.1's vertical video support shows that AI video creation models have begun to target the short-form content market in earnest. It is highly likely that other major AI image generation models will quickly introduce similar functions in the future.

Additionally, enhanced 'control' is expected to progress beyond prompt-based creation to more precise editing functions. If functions that allow detailed control of camera angles, lighting, character movements, etc. are added, AI tools can be fully integrated into professional video production workflows.

If Google integrates Veo with Google Cloud or YouTube Creator Tools, it is likely to have a significant impact on the entire content creation ecosystem. However, details such as the exact release schedule, API pricing policy, and regional availability have not yet been disclosed, so the timing of actual use will likely have to wait for a future announcement.