Meta boasts that its AI video can “outperform” competitors like OpenAI

• Meta just launched Movie Gen, an AI video generator designed to compete with OpenAI's Sora.
• Movie Gen can create videos with accompanying audio using a text prompt. It can also be edited by prompt.
• Meta later joined the video generation race as OpenAI and Google.

Meta released a new AI video generation tool on Friday, also the company's latest foray into the battle with OpenAI for AI supremacy.

“Today we are pleased to introduce Meta Movie Gen, our groundbreaking generative AI research for media, spanning modalities such as image, video and audio,” the company said in a press release. “Movie Gen outperforms similar models in the industry at these tasks when evaluated by humans.”

In its press release, Meta described Movie Gen as the “most advanced and comprehensive suite of storytelling models,” including video generation, audio generation, personalized video generation and video editing. The models were trained using publicly available and licensed data, the company said.

With a text prompt, Movie Gen can create videos up to 16 seconds long and 16 frames per second, thinking about “object movement, subject-object interactions, and camera movement.” Users can upload an image of themselves to incorporate into personalized videos, and Movie Gen can edit videos with text instructions from the user.

Meta's example video shows an underwater perspective of a baby hippo (Moo Deng reference, anyone?) happily swimming around in a calm water scene.

Another shows a koala on a surfboard and the accompanying prompt: “A fluffy koala bear is surfing. He has gray-white fur and a round nose. The surfboard is yellow. The koala bear holds on to the surfboard with his paws. The koala.” The bear’s facial expression is focused. The sun is shining.

With audio generation, users can create and extend “sound effects, background music or entire soundtracks” up to 45 seconds long, the press release says. An example clip of a snake Gliding through a forest area includes the Prompt: “Rustling leaves and breaking branches, with an orchestral piece of music.”

Meta is a bit late to audio and video generation, as top competitors like OpenAI and Google have already established a foothold in this space. OpenAi launched its Sora video generator in February, and Google followed with Veo in May.

However, Meta has given OpenAI a run for its money in the AI ​​arms race. Although OpenAI's ChatGPT was first to market and brought the company global exposure, recent iterations of Meta's Llama model have been well received. Many considered Llama 3.1, which came out in July, to be superior to OpenAI's GPT-4o, which came out shortly before.

According to Meta, its new “state-of-the-art models” outperform the competition in human A/B comparisons. When it comes to video generation, companies surveyed by Meta preferred Movie Gen over OpenAI Sora, the company's press release said. Meta hasn't published an A/B comparison with Google's Veo, which also features sound effects and music, but Meta said in a detailed accompanying research paper that it expects Google's video-to-audio generation models may be more limited in length are than those of Meta.

Meta, OpenAI and Google did not immediately respond to a request for comment.