Agentic planner
An LLM-grounded planner turns grasps into multi-step tasks.

Agentic skill API that sequences and recovers grasps into tasks.
A managed agentic API that sequences Dextra grasp skills into multi-step tasks, monitors execution through Tactyl's predictions, and autonomously re-plans and recovers when a grasp fails.
Built to be GPU-essential, NVIDIA-aligned, and production-ready.
An LLM-grounded planner turns grasps into multi-step tasks.
Detects slip or miss and re-plans a recovery grasp automatically.
Compose reusable manipulation skills into task graphs.
Reasoning is anchored in real Tactyl contact signals.
Call it from your stack — no scripting every contingency.
Design and monitor workflows visually.
| Type | Agentic skill/task API |
|---|---|
| Reasoning | NVIDIA NeMo LLM |
| Serving | Triton + NIM microservices |
| Scheduling | NVIDIA cuOpt |
| Edge | Jetson Thor grounded execution |
| Delivery | NVIDIA AI Enterprise |
Start free. Scale to a fleet. Pricing shown is indicative, pre-launch.
Cross-embodiment VLA manipulation foundation model + fine-tuning toolkit.
Explore Dextra →
Tactile-and-vision contact world model that predicts grasp outcomes.
Explore Tactyl →
On-robot edge runtime for low-latency closed-loop manipulation.
Explore Reflex →The manipulation layer teams ship with, not a research demo.
One policy runs across many grippers and hands — no per-robot rewrite.
Tactyl predicts slip and force so grasps react before they fail.
Reflex executes on Jetson at control-loop latency, with safety limits.
Every deployment feeds the shared model through the data flywheel.
Benchmarks, sim-to-real notes, and product updates. No hype, no spam.
Tell us your robot and task — we'll scope a pilot.