Tavus Unveils Griffin, First AI Model to Pass Video Turing Test with 48% Human Believability
Sign in✓ Signed in
Tavus has introduced Griffin, a video-to-video AI model that achieves unprecedented realism in face-to-face conversations, with nearly half of test participants unable to distinguish it from a human. The company announced Griffin on October 1, 2026, as a full-duplex conversational video system designed to mimic natural human interaction—seeing, hearing, speaking, and reacting simultaneously.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...