OpenAI’s 12-day “ship-mas” product rollout culminates with Sora, a powerful text-to-video AI model. Launched on Sora.com, it caters to existing ChatGPT subscribers in most regions (excluding Europe and the UK for now).
Read: Bandwidth Blog & Smile 90.4FM Tech Tuesday: New handheld PlayStation!
Sora offers a range of features: generate videos from text descriptions, animate still images, and even remix existing videos. The base ChatGPT Plus subscription allows 50 priority videos (720p, 5 seconds) with a watermark, while the $200/month Pro tier grants unlimited generations (1080p, 20 seconds), watermark removal, and simultaneous processing.
This isn’t entirely new – OpenAI teased Sora in February. However, a recent livestream by tech personality MKBHD confirmed the official launch and showcased some exciting capabilities. Users can craft videos by providing a sequence of text prompts, transform photos into moving scenes, and even refine existing AI-generated footage with further instructions. Additionally, a “remix” tool allows blending two scenes seamlessly.
OpenAI remains cautious regarding potential misuse. All generated content will have watermarks and C2PA metadata for identification. Users uploading media must confirm it adheres to strict guidelines, prohibiting content with minors, violence, or copyright infringement. OpenAI acknowledges the challenge of balancing creative freedom with responsible AI use and is open to user feedback.
While a free tier doesn’t exist, the public can still browse the AI-generated video feed created by the Sora community. This launch comes amidst controversy, with a group claiming to be alpha testers leaking the product in protest of alleged unpaid testing and promotion.
