Meta likely working on three image and video generation AI models: All detailed here

Meta is once again in the limelight, but this time not for the traditional upkeep of its social media platforms for engagement boost, but for developing three new AI image and video generation models.

As per reports making the rounds on the internet, Meta's purportedly upcoming AI models are being built by the Facebook maker's niche AI division, Superintelligence Labs, which it created earlier this year.

Of the tech juggernaut's three new AI models, one is said to be focused on image and video generation, to compete with offerings from Google and OpenAI.

Mango: Meta's latest AI image generation model

Citing the Wall Street Journal's insights into an internal Q&A with Alexandr Wang, head of Meta Superintelligence Labs (MSL), Gadgets360 noted that the image and video generation model, codenamed “Mango,” will likely be released by the middle of 2026.

Avocado: Meta's latest video generation model

Besides Mango, the other AI model Meta is working on is a text-based large language model (LLM) called “Avocado,” which is expected to focus on coding and share a similar release timeline.

Meta's new AI models signify an effort to keep up with tech giants like Google and OpenAI, which have explored coding-focused AI to attract enterprise clients.

The most intriguing part of MSL is that it has gone deep in developing world models, which enable AI systems to learn about the real world through visual input.

Meta's initiatives in image and video generation will directly rival Google’s Nano Banana and Veo, as well as OpenAI’s ChatGPT Image and Sora, developments that highlight its commitment to competing with established leaders in the AI industry.