Mistral Launches Robostral Navigate, an 8B Robot Navigation Model Using One Camera

Mistral AI has released Robostral Navigate, its first robotics model, an 8B-parameter system that guides robots through natural-language navigation tasks using nothing but a single RGB camera, no LiDAR or depth sensor required.

• 76.6% success rate on unseen R2R-CE navigation scenes, 9.7 points above the best single-camera baseline and 4.5 points above leading multi-sensor systems
• Trained entirely in simulation on about 400,000 trajectories across 6,000 scenes, then refined with online reinforcement learning
• Works across wheeled, legged, and flying robot platforms without hardware-specific retraining
• Navigates by pointing to a target location and orientation in the camera view rather than computing metric displacements

The release puts Mistral, alongside labs like Physical Intelligence and Figure, into the physical AI race that Meta and Google have also been pushing into this year. Access is currently limited to select partners in manufacturing, logistics, delivery, and hospitality rather than open weights.