Tribe v2 is an AI model developed by Meta that predicts how the human brain responds to visual, audio, and language inputs by simulating brain activity.
Meta Just Trained An AI On 700 Human Brains & Now It Can Predict How Yours Works
Meta's Tribe v2 AI can predict how your brain responds to images, sounds, and text, all without a single brain scan. Here is what this means for science and medicine.

- Meta unveils Tribe v2 AI predicting brain response to stimuli.
- Model uses high-resolution fMRI data from 700+ volunteers.
- Tribe v2 simulates brain activity without constant human testing.
Meta has launched Tribe v2, an artificial intelligence model designed to predict how the human brain responds to visual, audio, and language inputs. Built on high-resolution fMRI data from more than 700 healthy volunteers, the model can simulate brain activity without requiring human subjects in every experiment.
The move marks a significant step in using AI to better understand the brain, with potential implications for treating neurological conditions that affect hundreds of millions of people worldwide.
How Was Tribe v2 Developed?
Tribe v2 builds on Meta's Algonauts 2025 award-winning model, which was originally trained on low-resolution fMRI recordings from just four individuals. The newer version draws on a far larger dataset, with over 700 healthy volunteers exposed to a wide range of media, including images, podcasts, videos, and text.
ALSO READ: Meta Is About To Charge You Rs 775 For Using AI On Instagram And WhatsApp
According to Meta, Tribe v2 reliably predicts high-resolution fMRI brain activity and is capable of zero-shot predictions, meaning it can make accurate forecasts for new subjects, languages, and tasks it has not previously encountered. The model consistently outperforms standard modelling approaches used in the field.
The core idea behind the model is to create a digital representation of the human brain, allowing researchers to rapidly test hypotheses about how the brain functions without needing to involve human participants at every stage of an experiment.
What Does Meta Plan to Do With Tribe v2?
To support further research, Meta is releasing a research paper alongside model weights and code under a CC BY-NC license. A demo website has also been made available for anyone who wants to explore the model directly.
ALSO READ: This Forza Horizon 6 AI Driver Has Already Crashed More Cars Than You Have Ever Owned
The company says it hopes the release will help accelerate neuroscience research and open new paths in clinical practice. Understanding the brain more precisely could also benefit AI development, as neuroscientific principles can directly inform how these systems are built and improved.
Before You Go
Apple creates a new record in iPhone sales after launch of iPhone 16 | ABP Paisa Live
Frequently Asked Questions
What is Meta's Tribe v2?
How was Tribe v2 developed?
It was built using high-resolution fMRI data from over 700 healthy volunteers, significantly expanding on its predecessor's limited dataset.
What makes Tribe v2 unique?
Tribe v2 can make accurate 'zero-shot' predictions for new subjects and tasks, and it consistently outperforms standard modeling approaches.
What is Meta's goal in releasing Tribe v2?
Meta aims to accelerate neuroscience research and open new clinical paths by providing the model's weights and code for researchers to use and explore.


























