How to Try Open AI’s Sora Text-to-Video Generator

 

We have got ChatGPT, we have got DALL-E and now we have Sora from the labs of OpenAI. This Text-to-Video Generator is expected to have the same impact as ChatGPT did in 2022. How can you access it right now? Is Sora AI available to use? Let’s find out!

What is Sora by OpenAI?

Sora is an AI model made by OpenAI that can create minute-long videos using AI from the given text prompts. The videos are high-quality and look quite realistic. It can create videos with multiple characters, an object in motion, and whatever background the user wants. It can even make multiple shots in a single video.

The model also understands how things are expected to exist in the real world, so there is no need to give very detailed prompts. Still, there are some things it might not do right now. OpenAI has admitted that, but we expect it will change with time. Along with Text Inputs, you can prompt images in Sora to create your animation.

How to Access Sora AI Video Generator?

Update: OpenAI launched Sora on 09 December 2024 to the general public in the US and many other countries. The ‘turbo’ model was made available for ChatGPT Pro and Plus subscribers.

OpenAI announced the Sora Ai model on 15 February 2024 but it was not made available for the general public to use. Sora was only accessible by OpenAI’s Red Teamers. They have also granted access to some filmmakers and visual artists to get more feedback on how to make the product more useful for the general public.

The Red Teaming Network is a community of domain experts from different fields to try OpenAI’s products and assess any technical or safety concerns associated with them. These experts have experience in a particular domain and are passionate about improving AI safety.

In this particular case, they will be testing it with Red Teamers with experience in preventing the model from misinformation, bias, and hateful content. Note that the application to become a part of this network is closed. Along with this beta testing, they have implemented the same safety methods used by DALL-E 3.

However, OpenAI CEO Sam Altman has been asking users on X (formerly Twitter) to give him prompt ideas and he is sharing AI-generated videos with Sora. Here are some examples:

Prompt: “Two golden retrievers podcasting on top of a mountain”

Prompt: “A bicycle race on the ocean with different animals as athletes riding the bicycles with drone camera view”

Prompt: “An instructional cooking session for homemade gnocchi hosted by a grandmother social media influencer set in a rustic Tuscan country kitchen with cinematic lighting”

You can check some more examples from OpenAI’s Sora website, they have uploaded a lot of videos for different types of prompts.

Can you use OpenAI Sora?

If you are part of the Red Teaming Network, you may get access if they consider you to be an expert who may help them in making the AI better. Their DALL-E model also took a couple of months to be open for all the users. Most likely, it will be under their paid subscription only, but they may give some free credits as a trial.

There is still no official announcement made by the company on when it be unveiled for personal use, but we will update this news as as soon as we get new information.

Conclusion

Sora by OpenAI is not open to the public as of now, but it is going to be soon, in a few weeks or months. But there are some concerns about how the model is trained and if they have taken consent from people, whose videos are the dataset for the model. This might have some effect on OpenAI’s plan for this AI video generator.