Skip to content
Skip to main content
header search
Craftsman Practical Guide
Menu
sticky search

OpenAI launches Sora, an AI tool that generates videos from text

OpenAI has announced a new AI model that can create videos from text. Say hello to Sora.

 

Sora was introduced earlier this year in a preview, but now the AI-powered tool has graduated from beta, bringing a newer model called Sora Turbo.

OpenAI's Sora can generate videos based on text, images and videos

The video generator's primary feature is easy to use, it's very similar to how artificial intelligence tools can be used to generate images. You enter a text prompt, describing the content of the video, and Sora creates the video for you. It offers additional features, such as allowing you to import images and video files.

OpenAI Sora Video Editor

Sora Turbo can replace, remove and re-imagine, remix your videos based on your requirements. You can edit the content of your timeline, aka the Storyboard, by cutting, organizing the frames to create a personal timeline, looping videos, or combine two videos into a single clip. Sora can create videos based on a collection of photos. The AI offers presets with styles and filters that you can choose from. You can watch videos that were created by Sora at OpenAI's website.

The maximum length of a video that Sora can create is 20 seconds, and there are more limitations in how one can use the tool. Videos generated by Sora have a watermark. There are many precautions in place, users cannot use the service to create illegal, explicit, or violent content.  Sora is not available for free users, you will need to have a ChatGPT subscription. Let's talk about this.

OpenAI Sora pricing and limitations

OpenAI had announced a ChatGPT Pro subscription that costs $200 per month, and while the price seemed crazy, it kind of makes sense now. On the other hand, ChatGPT Plus only costs $20 a month. ChatGPT Plus users get 1000 credits for Sora, which offer up to 50 Priority videos, but does not allow access to Relaxed videos. ChatGPT Pro users have access to 500 Priority videos (10,000 credits), and unlimited Relaxed videos. Pro users can download videos without watermarks.

Hang on, what are Priority Videos and Relaxed Videos? What are Sora credits? It's like this, credits are sort of tokens or currency that you can spend for generating Priority videos. A 20-second video in 1080p square resolution costs 1000 credits, but that's for ChatGPT Pro users. ChatGPT Plus users can only create 720p videos which are 5 seconds long.

Relaxed videos do not consume credits, I guess you can call them free. The difference between Priority videos and Relaxed videos is kind of odd. Priority videos as the name suggests generates videos faster. When you run out of credits, Sora switches to Relaxed mode that queues up videos to be completed. Relaxed videos are only available for the higher-priced tier, ChatGPT Pro. Basically, this is a restriction in place to prevent the servers from getting hammered.

OpenAI Sora pricing and limitations

And no, Sora credits cannot be stacked up, they do not roll over to the next month. You either use the credits or they expire. There is a whole help page about all this, you can find the answers here.

Many people may feel that the price is too steep to create 20-second videos. I get it, you are probably thinking what are you going to use it for, besides as a screensaver or a personalized slideshow of some sort? For example, as a gamer and a bookworm, and I would like to experience what a particular fantasy world looks like in a video or art form, but is that worth $20? That depends on your spending power and needs. I think that tools like these might offer more value to artists, designers, etc.

Sora is not included with ChatGPT Team, Enterprise, or Edu. You need to be 18 years of age or older to access the AI. Sora is available for all ChatGPT subscribers, except those in the United Kingdom, Switzerland and the European Economic Area.

Anyway, if you're interested in trying Sora or learning about its capabilities, you may want to check out the official help section and tutorials on OpenAI's portal, or the OpenAI YouTube channel.

On a side note, Sundar Pichai, Google's CEO, said that the progress of AI tools have begun slowing down, and that the low-hanging fruit is gone. He's correct, AI tools like OpenAI's ChatGPT, Google Gemini, Microsoft Copilot, Apple Intelligence, etc., have evolved from being mere chatbots to generating code, writing letters, poems, stories, summarizing text content on pages, generating images, and creating videos.  There is word already about ads in AI apps, but the next real big step in AI's evolution will take time. But, do we need to rush the advancements of AI? Would it not be better to improve the current AI models?

AI moderation needs to be a top priority. Most, if not all, AI bots have many flaws. The biggest problem is, of course, combating fake news and incorrect information, but there are other concerns such as misusing of data, or potentially causing harm to others. Big Tech needs to take a deep breath, analyze the situation, and ensure that AI tools are made safer to use.

What do you think about Sora? Cool? Scary? Useful? Meh?

More exciting content