FLUX-mimic-x: 7B video-action model tops RoboLab leaderboard at 51.4%, zero-shot across embodiments
robrombach · x · 2026-11-16
Mimic Robotics previewed FLUX-mimic-x, a frontier Video-Action Model that takes the #1 spot on the RoboLab benchmark among real-time "system 1" models, with zero-shot generalization across grippers and dexterous hands.
Key facts:
- 7B parameters, built on the updated mimic-video recipe and trained on top of FLUX 3 by Black Forest Labs
- Never trained on simulation data or any RoboLab-embodiment data beyond the original DROID dataset
- Deploys on any robot and task with minimal adaptation, running on-device in real time
- RoboLab task success rates: FLUX-mimic-x (DROID FT) leads at 51.4%, ahead of GROOT 2 (47.4%), Praxis-1 (46.2%), FLUX 3 Action (42.9%), Cosmos 3 Nano (36.8%), π0.5 VLA (28.0%), Dream Zero (25.7%), and GROOT N1.6 (7.2%)
Related event: FLUX-mimic-x: 7B Video-Action Model Tops RoboLab Realtime Leaderboard(2 posts)→
More from Embodied
- Tesla to invest €3B in Giga Berlin expansion, targeting up to 1M Cybercabs by 2029 — XFreeze · 2026-11-16
- Lightwheel launches RoboFinals RL-Env: RL training envs from real industrial tasks with calibrated physics — chris_j_paxton · 2026-11-16
- Day 11 of a DIY robot arm: joint reworked to ride on a bearing with gear drive — _Stocko_ · 2026-11-16
- Robot Learning Seminar: JHU's Homanga Bharadhwaj on predicting human behavior for dexterous manipulation — mangahomanga · 2026-11-16
- Humanoid robot works hotel night shift in Shanghai, 165 real laundry trials over 57 hours — CyberRobooo · 2026-11-16
- Robot cup-handling demo goes viral: less gymnastics, more arguing with gravity — TansuYegen · 2026-11-16