Meta expected to release its Avocado AI model in Q1

Meta’s AI kitchen is cooking

Have you noticed AI popping up everywhere, from chatbots to weirdly good photo filters? The tech giants are in a heated race to build the smartest systems. Meta, the company behind Facebook and Instagram, is quietly working on its biggest recipe yet.

These aren’t small updates. They represent Meta’s ambitious attempt to catch up to leaders like OpenAI’s ChatGPT and Google’s Gemini. The company is targeting the first half of 2026 to serve these new models to the public, hoping they’ll be a game-changer.

Meet Avocado, the text whiz

First in the duo is Avocado, a powerful new language model. Think of it as a supercharged brain for understanding and generating text. Its special focus is expected to be on improving coding and complex reasoning skills.

This makes Avocado a direct competitor to models that help programmers write software. If successful, it could become a vital tool for developers and a core part of Meta’s own apps and services.

And mango, the creative visionary

The second model, Mango, is all about visuals. It’s being designed as a next-generation image and video generator. The goal is for it to create high-quality, coherent video clips from simple text descriptions.

Beyond just making pretty pictures, Mango aims to understand how the visual world works. This means potentially generating videos where objects move and interact in physically believable ways.

A major strategy flip

Here’s the surprising twist: these new models might not be free for all. Meta publicly released Llama family models and provided access to model weights, but those releases were accompanied by a Meta community license and acceptable use restrictions rather than an unrestricted open source license. For Avocado and Mango, the company is considering keeping them proprietary.

This represents a strategic shift: Meta appears to be weighing a more closed approach for its next-generation models, which would be a noticeable change from how some earlier Llama models were distributed.

New chef in the AI kitchen

To helm this crucial project, Meta brought in a new leader, Alexandr Wang. Meta invested roughly $14 billion in Scale AI and recruited its founder, Alexandr Wang, to join Meta as the leader of the new Meta Superintelligence Labs

Wang now oversees the elite team building Avocado and Mango. His hiring signals Meta’s desire for fresh, aggressive leadership to accelerate its AI development and adopt fast-paced methods.

Internal culture clash

This new leadership has reportedly changed how Meta’s AI teams work. The focus is now on speed, with a demo, don’t memo mantra encouraging rapid prototyping over lengthy planning documents.

There has been notable turnover and reorganization; Yann LeCun has announced plans to depart Meta to start a new AI company, and other senior researchers have also moved on.

Pressure to catch up

Why is Meta making such a dramatic push? Simply put, they feel they’re falling behind. While their social media ads are hugely profitable, their standalone AI products haven’t captured the public’s imagination.

Releases from competitors have been coming fast and impressing users. Avocado and Mango are the company’s best to regain its seat at the very top table of AI innovation.

The immense cost of the race

Staying in this AI race requires unbelievable spending. Meta has raised its 2025 budget for capital expenditures to a staggering $70-$72 billion. A huge portion of that is dedicated to training these massive AI models.

This spending is a gamble that these future AI systems will eventually pay for themselves. Wall Street is watching closely, looking for a return on this enormous investment.

Where you might see it first

So, where could you actually use an Avocado or a Mango? Likely, right inside the apps you already have. Meta could integrate Avocado’s smarts into the Meta AI assistant in Facebook, Instagram, and WhatsApp.

Mango’s video skills could supercharge features for creators, making it easier to produce engaging content. Imagine AI-powered editing tools in Instagram Reels or new ways for advertisers to quickly make video ads.

The bigger ambition, world models

Beyond today’s tools, Meta’s research points to a grander idea. They are exploring world models, AI that learns from visual input to understand how environments work. This is about more than creating a video.

Such technology is a step toward the long-term dream of more general, adaptable AI that can plan and perform complex tasks across different situations.

The open source question mark

Meta’s potential move away from open-source AI raises big questions. Releasing models like Llama for free allowed researchers and small companies to innovate, speeding up overall progress in the field.

The company has cited safety and competitive concerns. The debate continues over how the fastest AI advancement happens, through open collaboration or private competition.

A defining moment for Meta

The success or failure of Avocado and Mango is about more than two products. It’s a test of Meta’s entire AI reorganization and its massive investment. Can this new strategy deliver a true breakthrough?

The outcome will shape perceptions of whether Meta can be a primary driver of AI’s future. The first half of 2026 will be a critical reveal.

Want to see where else all those billions are going? Check out how Meta keeps paying billions in fines.

Why this matters to you

You might not care about corporate strategy, but this AI race directly influences the tools you use daily. The competition pushes companies to make AI assistants smarter and creative tools more powerful.

As these models develop, questions about their responsible use and how they shape the information we see will grow more important. Understanding these shifts helps us all navigate our increasingly AI-powered world.

Curious about another huge financial move they just made? Take a peek behind the curtain at why Meta backs away from $ the $600B Trump deal.

Which of these AI projects has you more intrigued, Avocado for coding or Mango for video? Share your thoughts below and like this post if you enjoyed the deep dive.

This slideshow was made with AI assistance and human editing.

Don’t forget to follow us for more exclusive content right here on MSN.

Read More From This Brand:

• Apple loses key AI search executive to Meta
• Meta strikes massive cloud computing deal with Google
• Meta and Arm strike a partnership to expand global AI capabilities