Editorial illustration for Meta AI lab to debut ‘Avocado’ text model and ‘Mango’ vision model in Q1
Meta's Mango and Avocado AI Models Set to Disrupt
Meta AI lab to debut ‘Avocado’ text model and ‘Mango’ vision model in Q1
Meta plans to ship two new AI models this quarter, internally codenamed Avocado and Mango. This is the company's opening move after a year of costly restructuring, new labs, and talent raids. It’s a catch-up play.
The text model Avocado and vision model Mango are the first major deliverables from that push. CEO Mark Zuckerberg wants Meta to lead, not just participate. But CTO Andrew Bosworth is already tempering expectations. The real work, he says, begins when the training runs stop.
According to WSJ, in December, Meta was developing a text-based AI model codenamed 'Avocado', expected to launch in the first quarter, alongside an image- and video-focused model codenamed 'Mango'. Bosworth did not confirm which of these models had been delivered internally. Meta's progress is being closely watched after chief executive Mark Zuckerberg moved to overhaul the company's AI leadership, set up a new lab and aggressively recruit top talent with lucrative compensation packages, as he seeks to position Meta at the forefront of AI development.
Bosworth cautioned that building usable AI systems involves far more than training models alone. "There's a tremendous amount of work to do post-training to actually deliver the model in a way that's usable internally and by consumers," he said. He described 2025 as a "tremendously chaotic year" for Meta, marked by rapid infrastructure build-out, expanded computing capacity and efforts to secure sufficient power to support its AI ambitions.
Bosworth’s “tremendously chaotic” warning for 2025 is telling. The race has shifted. It’s no longer purely about model intelligence; it’s about building power grids, data centers, and product pipelines at scale.
Meta is betting it can manage that industrial chaos better than its rivals. These fruit-named models are just the start.
Related Reading
Browse all Meta news →
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
Related Topic WhatsApp launches Meta AI Incognito Chat, cuts latency for privacy
AI Tools & Apps
Related Topic Meta's DreamGym boosts AI agent success by 30% over baseline methods
LLMs & Generative AI
From Archives Telangana Launches Aikam AI Innovation Hub at Davos, Calls for Global Partners
From Archives TrueFoundry launches TrueFailover to auto‑reroute AI traffic on model outages
From Archives Google’s internal RL enables metacontroller to learn abstractions on frozen models
Industry Applications
From Archives Zuckerberg Unveils Meta Compute to Build Global AI Infrastructure
Common Questions Answered
What are the two AI models Meta is developing under its Superintelligence Labs?
Meta is developing two generative AI models: 'Mango', which focuses on image and video generation, and 'Avocado', a text and code-based model. Both models are expected to launch in the first half of 2026 and are part of Meta's strategy to catch up with competitors like OpenAI and Google in the AI race.
Who is leading the development of Meta's new AI models?
The AI models are being developed under Meta Superintelligence Labs (MSL), which is led by Alexandr Wang, co-founder of Scale AI. The project also involves Meta's Chief Product Officer Chris Cox, with the goal of creating AI systems that can reason, plan, and act without needing training for every specific scenario.
What are the specific capabilities of the 'Avocado' and 'Mango' AI models?
'Avocado' is designed to excel in text generation, coding, and logical reasoning, with a priority on improving coding capabilities and performing advanced text analysis. 'Mango' is focused on generating high-fidelity images and videos, aiming to compete with other top-tier visual content creation engines in terms of quality, realism, and creative control.
Further Reading
- Meta is developing a new image and video model for a 2026 release — TechCrunch
- Meta's Superintelligence Labs Hits First Breakthrough AI Models — eWeek
- Meta's new AI team delivered first key models — The News International
- Meta's new AI roadmap: Mango and Avocado models — ContentGrip
- Meta readies next-generation Mango and Avocado AI models for 2026 launch — MLQ.ai
