Open AI Releases Sora The AI Video Generating Model. Here's What You Can Do With It.
News & Politics
Introduction
OpenAI's long-rumored and much-anticipated video-generating AI model, Sora, has officially launched and is now accessible at sora.com. This innovative tool empowers users to create videos with a resolution of up to 1080p, lasting up to 20 seconds in various aspect ratios, including widescreen, vertical, or square.
Key Features
Users can either upload their own assets to remix and blend existing content or generate entirely new videos from textual descriptions. Sora's capabilities are included as part of a Plus account at no additional cost for OpenAI users. Each account allows for the generation of up to 50 videos at 480p resolution or slightly fewer at 720p per month. However, it is important to note that Sora is currently unavailable in the EU and China.
OpenAI has also introduced a Pro Plan for those looking to leverage Sora more extensively, offering 10 times more usage, higher resolutions, and longer video durations. Pricing tailored for various user types is expected to launch early next year.
Limitations and Safeguards
While the version of Sora currently deployed boasts impressive features, it does come with some limitations. Users may find that it often generates unrealistic physics and struggles with complex actions over extended durations. Nevertheless, Sora Turbo is significantly faster than the model released earlier this year.
OpenAI is aware of these challenges and is committed to making their technology affordable and accessible. They are also proactively encouraging users to responsibly explore video generation technology. Each Sora-generated video is tagged with C2P metadata to ensure transparency and can be verified for its origin. To enhance user safety, Sora includes visible watermarks by default and features an internal search tool to help verify the source of content.
Accessibility
According to the Los Angeles Times, Sora will be accessible to individuals aged 18 or older in regions where ChatGPT is available, excluding the United Kingdom, Switzerland, and countries in the European Economic Area. OpenAI is actively working on making Sora available in these locations as well.
Looking ahead, OpenAI is also preparing to release a free version of Sora. Suki Mansour, the program lead for OpenAI’s artist program, emphasized that Sora is designed to enable artists to achieve ambitious projects without the burden of expensive resources, thereby democratizing video creation.
How Sora Works
Sora functions similarly to OpenAI's image-generating AI tool, DALL·E. Users simply type in a description of the desired scene, and Sora produces a high-definition video clip in response. Notably, Sora can also create video clips inspired by still images and fill in missing frames of existing videos.
With the introduction of Sora, users can look forward to a surge of creativity in video content across social media platforms.
Keywords
- OpenAI
- Sora
- AI video generation
- 1080p resolution
- Video creation
- C2P metadata
- Pro Plan
- Creative collaborator
FAQ
What is Sora?
Sora is an AI model developed by OpenAI that generates videos based on user-provided text descriptions or input assets.
What resolutions can Sora generate?
Sora can produce videos up to 1080p resolution.
How long can the videos be?
Videos generated by Sora can be up to 20 seconds long.
Is Sora available in all countries?
No, Sora is currently unavailable in the EU and China. Access is also restricted in the UK and Switzerland.
What options are available for users who want more from Sora?
A Pro Plan is available that offers increased usage, higher resolutions, and longer durations for users who need more access to Sora.
What safeguards does Sora have in place?
Sora implements C2P metadata for transparency, visible watermarks by default, and an internal search tool to verify content origin.
