VLA models
Robostral Navigate
France
Mistral AI's entry into physical AI (announced 8 July 2026): an 8B model steering wheeled, legged and flying robots from a single RGB camera and plain-language instructions, trained entirely in simulation. Availability and licensing had not been announced at launch.
OrganizationMistral AI
CountryFR
Release2026-07
Params (B)8
Open weightsno
Architecturepointing-based navigation model: grounding-specialized Mistral VLM predicts target image coordinates and desired orientation in the current camera view, with local displacement fallback
Modalitiesvision, language, action
Embodimentsmobile, legged, aerial, cross-embodiment
Training data~400,000 simulated trajectories across 6,000 scenes, generated entirely in-house without open-source VLMs; further improved with online reinforcement learning (+3.2 pts reported)
Sources: Mistral AI (2026-07-08)The Decoder (2026-07-08)Bloomberg (2026-07-08)
← Back to comparator · VLA models