OpenAI’s Sora Is Here: Everything You Need to Know About the Video Generator | by hejrene | Rene | Dec, 2024 | Medium

New AI tool

OpenAI’s Sora Is Here: Everything You Need to Know About the Video Generator

I’m amazed by how OpenAI’s Sora brings stunning creativity to video, even for someone like me who’s not a “video person.”

Hej Ai friends and followers.

Today is the day! OpenAI has just officially released Sora, their video-generating tool, to the public. It was announced a couple of month ago and I am sure no one expected it to launch this early?

What is it?

If you haven’t been living under a rock, you should already know that Sora is the video equivalent of DALL-E, OpenAI’s image-creation tool. Sora was first announced in February and impressed with video clips that, in my opinion, were far superior than to the competition.

So is it any good today? Well I guess that is for everyone to decide, but to quote early reviews there are some things which are remarkable and some things that need improvement. Do you remember the early days of image generators? Back in the day when Midjourney was still in its infancy? They released new models in rapid succession and with each new updates the outputs become increasingly realistic.

I believe the same will happen with Sora, except that a lot of the early footage already looks pretty realistic. But there are other problems which give away that the videos were made with AI (and maybe this is a good thing right now in these uncertain times?).

For instance physics seem to be a real challenge. And how objects behave in space. Sometimes, objects even appear out of nowhere or move through each other. All things that can and will be resolved I guess. And I can only imagine what Sora will be capable of in 6 months or a year from now!

Features

What I really like about this first version is that I comes with a bunch of sorting and management functions. You can create folders, upload images, favorite your own or other creations and get inspired by others.

You can also remix existing videos and add content by prompting just like you would in the new editor in Midjourney for instance. And Sora features a storyboard, which allows you to connect several prompts for a consistent video. That is really cool and helps to create longer scenes, because each clip is just up to 20 seconds long (depending on the subscription — see below).

There are other features too like blending videos together or creating loops. The feature set is rich and it will take some time to explore everything.

Safety

But so far the results are really impressive. Is it changing the game? I am not sure, I guess it will take time. But I really appreciate and respect OpenAI’s safety precautions. Just like in their other products, Sora will reject prompts that are against their rules (ie. extreme violence, celebrity likenesses or IPs).

Availability

Sora will be available almost everywhere, except certain countries and of course Europe and the UK. Frankly it is frustrating how the EU has become an AI black hole. No Meta AI, no Apple Intelligence, no Sora, it took month to bring advanced voice mode to Europe. Now don’t get me wrong, I am all for data protection, but there is a fine line between genuine data protection and lagging behind technologically. And right now, Europe is on track to loose influence and falling behind.

Price

Pricing is a real surprise.

All Plus subscribers are getting 50 free creations each month (up to 720p and 5 seconds in duration). While that is not enough to create a feature length movie, that’s again not the intention right now.

And Pro subscribers are getting unlimited creations in a relaxed mode and 500 priority videos per month (up to 1080p and lasting up to 20 seconds). That is really fine with me.

Will I Use Sora?

I am pretty sure I will play with Sora once the excitement has died down. I have never been a video person though. I have the greatest camera in my pocket (an iPhone 16 Pro Max), but I just take photos.

That said, the creativity showcased on the Sora website is inspiring. The video clips are beautiful, and I’m all for anything aesthetically pleasing. I cannot wait to turn some of my Midjourney creations into video clips.

But overall I will stick with generative image tools like Midjourney for the time being.

⭐️About me: René is a writer on Medium, a formally trained web designer and graphics artist, a long-term Apple user (20+ years), and an AI enthusiast who switched fields and now works in neurology while studying psychology.

Hej there! Can I ask you a favour (it will really help me out to grow this blog)? If you find this article insightful, follow me please and clap 50 times. Thanks for reading.