EmbodiedHub LogoEmbodiedHub
Back to Search
τ

τ₀-VLA

Model

A hierarchical robot foundation model with world-model-guided test-time computation, developed by Shanghai Innovation Institute. The low-level vision-language-action policy is released as an Apache-2.0 pretrained checkpoint built on a Qwen3.5-2B backbone.

Details

Updated:8/20/2026
open sourcetrue
release date2026-08-17
github urlhttps://github.com/sii-research/tau-0-vla
paper urlhttps://arxiv.org/abs/2608.16885
model familyVLA
huggingface urlhttps://huggingface.co/sii-research/tau-0-vla

Tags

robot-foundation-modelVLAworld-modelopen-source

Relationships

No relationships found.

Sources

τ₀-VLA model card on Hugging Face
website
Visit
τ₀-VLA: A Hierarchical Robot Foundation Model with World-Model-Guided Test-Time Computation
website
Visit
τ₀-VLA official repository
website
Visit

Appears In

No related landscapes.