Fired OpenAI Finally, Sora, an AI video generation model, and in February, the company rolled out a beta version of Sora For select individuals, now, it has released a variant of the model called Sora Turbo, and Sora can create 1080p videos which can be up to 20 seconds long.
The AI model has been deployed on a standalone platform and is currently available as a website. Sora is currently only available to paid ChatGPT subscribers with specific rate limits.
Sora AI video generation model from OpenAI
In a blog post, the AI company announced the launch of Sora and detailed the model’s capabilities. Sora was first unveiled earlier this year, and the model has been repeatedly delayed. The company stated that the reason behind the delay was to enhance the safety and privacy standards of the model.
However, after a delay of nearly nine months, OpenAI has launched Sora as a standalone platform that can be accessed here. It is currently only available to ChatGPT Plus and Pro subscribers. Those without a subscription cannot create a new account on the site currently. Meanwhile, users are limited to Plus gets 50 480p videos or fewer 720p videos each month.
Subscription will be available ChatGPT Prowhich was recently introduced at $200 per month, allows users to create videos with “10x more usage, higher resolution, and longer durations,” however, just like “fewer videos,” the company did not specify what the higher resolution and longer durations might entail.
Sora can currently create videos in wide, vertical, and square aspect ratios. Users can also upload their own videos and images to stretch, remix, and combine content into created videos. The AI model also allows for videos to be created from scratch using text prompts. In addition, the storyboard interface allows users to assign specific inputs to each frame.
In terms of technical details, OpenAI explained that Sora is a diffusion model, where the AI has forward vision of many frames at once to maintain content consistency over a 20-second period. The AI model uses a transformer architecture, and takes the retranslation technology from DALL-E. 3.
OpenAI also highlighted details regarding the model’s data. The company claimed to have obtained a wide range of data from the public domain, through its data partnerships, and data from people working with the model. The public data was said to have been collected from machine learning datasets and searches. On the web.
The company also partnered with Shutterstock Pond5 and commissioned datasets to generate proprietary data for the AI model, and finally, Sora data was also collected from AI coaches, red team members, and employees.
To reduce the risks associated with a realistic AI video creation model, she adds OpenAI Both the visual watermark as well as the metadata are in accordance with the standards set by the Content Provenance and Authenticity Alliance (C2PA), and the company also claimed to have added protection in the form for media uploads involving people.
OpenAI introduces new ChatGPT Pro subscription, releases o1 AI model
The AI company also said that Sora will be banned from creating videos containing harmful forms of abuse such as pedophilia and deepfakes. Additionally, the number of uploads people can do will be limited at launch.
OpenAI, ChatGPT Plus, Sora Turbo, Videos, Artificial Intelligence
ظهرت في الأصل على www.youm7.com