Skip to content

OpenAI Unveils Sora 2: A Leap in Text-to-Video Generation

Sora 2 brings text-to-video generation to a new level. With improved quality and exciting features, it's free to use now, but OpenAI warns of potential risks and emphasizes safety measures.

In this image we can see a website. There are videos and we can see text.
In this image we can see a website. There are videos and we can see text.

OpenAI Unveils Sora 2: A Leap in Text-to-Video Generation

OpenAI has unveiled Sora 2, the latest iteration of its text-to-video model, marking a significant leap from its predecessor. The new model, presented in early October 2025, promises enhanced physical accuracy and photorealistic quality. It's initially free to use, with a premium tier available for ChatGPT Pro subscribers.

Sora 2 offers improved consistency, avoiding distortions in generated videos. It supports synchronized dialogues, sound effects, and background noises. The 'Cameos' feature allows users to integrate their own video and audio into Sora-generated scenes. The model is available through the new Sora iOS app and via sora.com, initially in the US and Canada through an invitation system.

OpenAI acknowledges potential risks such as misinformation and addictive potential. It has implemented safety measures like filters and moderation to mitigate these concerns. While Sora 2's resolution is limited to 1920 × 1080 pixels, OpenAI is committed to improving its capabilities over time.

Sora 2 is now available, offering enhanced features and improved quality. It's free to use initially, with a premium tier for ChatGPT Pro subscribers. OpenAI's focus on safety and continuous improvement promises an exciting future for text-to-video generation.

Read also:

Latest