Meta competes with OpenAI’s text-to-video generator Sora with its own AI model: Meta Movie Gen.
Meta introduces a new generative AI model that allows users to create videos based on a text prompt. Movie Gen can additionally edit videos and generate audio. In doing so, the company competes with OpenAI’s Sora model, among others, which can also convert text to video. Meta’s model is not yet publicly available. Meta says it is awaiting feedback from video creators.
Meta Movie Gen
Meta presents its new AI model Meta Movie Gen. This model is capable of performing several tasks. Based on a text prompt, you can have videos and audio generated. You can have an audio created for a video you enter, or a standalone audio. Moreover, it is possible to edit an existing video based on text prompts, or even personalize it.
read also
Meta introduces text-to-video generator Movie Gen
Still, there are some limitations associated with the model. Videos can be a maximum of 16 seconds, with a frame rate of 16fps and a maximum of 768 pixels. Audio files have a maximum of 45 seconds. Meta shares a paper on its Web site with more details on how the model was created.
Movie Gen appears to be a response to OpenAI’s Sora model, which is also capable of generating videos based on text prompts. Meta’s model is not currently publicly available. Meta says it wants to await feedback from video creators. Moreover, according to Meta, the model would be trained on datasets with “licensed and publicly available data.