A few days ago, the founder and CEO of OpenAI, the artificial intelligence company that presented ChatGPT at the end of 2022, announced on his X (formerly Twitter) profile the launch of a new platform that represents an important new step in the use of generative artificial intelligence. This is Sora, a tool that the general public cannot yet use, but the company assures that it has the capacity to create high-quality videos from text commands. What is striking is the great wealth of details that the machine can invent and the possibility of having several characters moving in clips of up to one minute.

To demonstrate this, OpenAI itself has published several videos that really draw attention due to the difficulty in differentiating whether the characters that appear are real pardons or created by artificial intelligence. Sora is currently only available to a series of visual artists, designers and filmmakers in a testing phase, to help the company improve the model and make it more useful for professionals when it is officially launched.

But there are already other tools on the market that work with generative AI for video creation. One of the best known is Runway, which last November launched its most powerful model, Gen-2. On this platform you can upload a static photo and the AI ​​will begin to animate it. It is also possible to write what you want with text and let Runway create a video. In testing, we asked for a cat running across the field and the program created a 4-second clip with a very real feline in the field (see photo). The free version allows you to carry out three tests, if not, you have to pay from 12 to 78 euros per month, depending on the amount of content you want to make.

Another very striking platform for how well it is capable of generating video from text is Pika. Like the previous one, it is as simple as writing a command about what you want to see on video and Pika will create it in a matter of a minute or two. In the free version you can only try 3-second clips and are limited to three videos, but with the paid version, for $28 per month, there are unlimited videos. In principle, in the tests we have done this tool does not work badly, but Runway offers more real content.

Another different concept, but one that is also based on the creation of videos using AI, is the possibility of creating avatars that give a certain speech. This is what platforms like HeyGen do. This tool allows you to choose a character, male or female, a language and a tone of speech, more formal or informal. Once this is done, you have to write a text that the chosen avatar will read in the indicated tone, and most strikingly, moving their lips as if they were really speaking. It’s so well done it’s scary.

As with the previous programs, it is possible to try and experiment for free, thanks to a series of credits that HeyGen grants when creating a profile, but to be able to create long videos or access all the tools it is necessary to pay, about 48 euros per month. if it is for personal use or 144 if it is for a company.

The same function as HeyGen is what Veed.Io has. This program also allows you to choose an avatar and the language in which you want it to speak, but it has many more options than the previous one, even incorporating different accents depending on the regions in which that language is spoken. In addition to the fact that it also synchronizes very well between the avatar and the movement of the lips or body to read the message, it has a series of very useful extra tools to, for example, add voice translations to several languages ​​or include subtitles . The free plan allows you to experiment a lot with Veed.IO, but to be able to make long videos you need to pay from 13 to 58 euros per month, depending on the plan you choose.

Another company that is experimenting with videos and AI but has not yet launched its tool is Meta. Mark Zuckerberg has been announcing some of the advances he is making in this field for some time and video generation with AI seems to be one of the most interesting to the tycoon. Emu Video was presented in November and although it is not yet possible to test it, according to Meta it will allow you to generate realistic and quality images and videos directly from a text command and directly in group chats or company applications.