OpenAI officially releases Sora, its text-to-video tool

(CLO) On December 10, OpenAI officially launched Sora, an advanced artificial intelligence (AI) model that allows creating videos from text, exclusively for ChatGPT Plus and Pro users.

The tool marks a major step forward in expanding OpenAI's multimodal AI technologies, competing directly with similar products from Meta, Google, and Stability AI.

OpenAI logo. Photo: REUTERS/Dado Ruvic

Sora supports creating videos with a resolution of up to 1080p, a maximum duration of 20 seconds, and can choose screen formats such as landscape, portrait, or square. Users can enter detailed text to create videos or upload personal images and videos for editing. In particular, the "Remix" feature allows changing the content of an existing video, while "Storyboard" helps combine text segments into a seamless video.

Sora is designed with safeguards to prevent abuse, such as prohibiting the creation of content that is violent, infringes on copyright, or involves child exploitation and deepfakes. The tool is currently unavailable in the EU, Switzerland, and the UK due to regulatory concerns, but has been released in many other regions.

While Sora allows for creative footage, the technology still has some limitations, such as inaccurate physical motion simulation, confusion in handling complex scenes, or creating unwanted objects in the video.

OpenAI plans to offer affordable pricing plans for a wide range of users as early as 2025. At the same time, the company is actively collaborating with experts to ensure safety when using this new technology.

Hong Hanh (according to Reuters, CNA)

Source: https://www.congluan.vn/openai-chinh-thuc-phat-hanh-sora-cong-cu-chuyen-van-ban-thanh-video-post324875.html