Generative AI tools are exploding, allowing users to create sharp articles, images, music, and videos from descriptive text.
The race between the tech giants is getting fiercer. At the end of May, Google introduced Veo 3 with the ability to create lip-synced videos and automatic voiceovers.
Not to be outdone, OpenAI released Sora 2, an upgraded version of the tool launched in February 2024.
Sora 2 stands out thanks to its accurate simulation of the laws of physics, making moving and interactive scenes come to life. In particular, the maximum video length is 10 seconds, slightly longer than Veo 3 (8 seconds).
Videos generated by Sora 2 in different styles (Video: OpenAI).
In addition to image rendering capabilities, Sora 2 also integrates background sound creation, character dialogue and supports many languages, including Vietnamese.
Another highlight is the Cameo feature: simply upload a short video recording your face and voice, and users can be "transformed" directly into the AI-generated video.

From sample videos of users, Sora 2 can combine them into AI-generated videos (Photo: OpenAI).
Sam Altman, CEO of OpenAI, calls Sora 2 “the most powerful imagination building tool ever created,” hoping to help people create entertaining videos, advertisements, or artistic ideas.
However, the face-injection feature also raises concerns that it could be misused to create misleading or offensive content. OpenAI said it will apply strict moderation and label “Sora” to identify AI-generated videos.
CEO Sam Altman is superimposed over videos generated by Sora 2 (Video: OpenAI).
Sora 2 is currently in limited testing on iOS in the US and Canada, with plans to expand to Android and other countries after the testing phase.
Source: https://dantri.com.vn/cong-nghe/video-va-am-thanh-an-tuong-tu-cong-cu-ai-moi-cua-openai-20251002125910913.htm
Comment (0)