Nvidia's TacitVLA matches CoT performance at up to 3.4x lower latency
arankomatsuzaki · x · 2026-10-14
Nvidia presents TacitVLA: the model still generates reasoning text (CoT) but masks it from its action expert, so actions don't wait for the text to finish. It matches the performance of its CoT counterpart at up to 3.4x lower latency. Project page and paper are available.
Related event: Nvidia Unveils TacitVLA, Cutting VLA Latency 3.4x with Tacit Reasoning(2 posts)→
More from Embodied
- FLUX 3 Action Tops RoboLab-120 at 42.9%, Bringing Video Pretraining to Robotics — DeepLearningAI · 2026-10-15
- DeployRobo launches as the generalist robot deployment company, backed by Y Combinator — ycombinator · 2026-10-14
- Robot foundation-model startups are all opening APIs — the model layer was never the moat — chris_j_paxton · 2026-10-14
- Metal Arm: a $2,499 6-axis CNC robot arm that ships in a week, built on LeRobot — chris_j_paxton · 2026-10-14
- NovaPlan (CoRL 2026): zero-shot long-horizon robot manipulation with no demonstrations — chris_j_paxton · 2026-10-14
- Ukrainian FPV Drone Sim Leaves Early Access, Players Generate Training Data — generativist · 2026-10-14