Skip to main content

Facialabuse-gaia-3 New! Jun 2026

: The model utilizes a new video tokenizer that is twice the size of the version used in GAIA-2. This allows it to capture: Safety-critical structures

. This version is optimized for Portuguese-language tasks and standardized exams (like ENEM), but it is distinct from the Wayve GAIA-3 autonomous driving model. Facialabuse-gaia-3