Motus: A Unified Latent Action World Model
Original research
MODEL PROFILE · Hybrid WAM / VLA
Motus
Combines understanding, video generation and action experts in one multi-purpose architecture.
Research organisationBi and collaborators
Research approachHybrid
Published year2025
BackboneMixture of Transformers
Overview
Motus uses an expert mixture to support multiple modes: world modelling, video generation, vision-language-action, inverse dynamics and joint video-action prediction. It is a hybrid rather than a single fixed prediction pipeline.
Source record
Core research identity and broad method were source-checked as of 2026-11-03. Independent benchmarking, commercial availability, weights and licensing have not been assessed.