gr00t-n1.7

Overview

Highest success rate of all 7 archs on LIBERO-Object (98.0%, RTX 3060), but same ~6GB footprint problem as N1.6 — does not fit the 8GB target hardware tier at all.

Architecture

Vision backbone

Qwen3-VL ViT

Language backbone

Qwen3-VL

Action head

AltVL DiT + self-attention (diffusion)

Action steps per chunk

16

Solver steps

4

Camera views

2 — video.image, video.wrist_image

Hardware fit

Benchmarks