# ModelScope says open-weight Qwen matches GPT-6 Astra on robot-use benchmark

> Original title: With uni-robot-agent, open-weight Qwen shows robot-use capabilities comparable to the published Astra–RoboProbe results on RoboDojo. The ...

- Source: ModelScope
- Published: 2026-11-03T15:11:12.000Z
- Bad Influence: https://badinfluence.news/items/qmqplplzmohni3fbq6rdlbgks
- Original: https://x.com/ModelScope2022/status/2117635714313802099

## Summary

ModelScope says its uni-robot-agent harness, running Qwen3.8-Flash-Next, scored 31.39 with 25.00% success on RoboDojo without privileged simulator inputs. That compares with 28.97 and 22.48% for the published GPT-6 Astra–RoboProbe configuration. Across 19 physical-robot tasks, the zero-shot Qwen agent scored 49.50/100, ahead of fine-tuned π0.5 at 47.75 and zero-shot Dream-1 at 35.50, according to the post.

## Post

With uni-robot-agent, open-weight Qwen shows robot-use capabilities comparable to the published Astra–RoboProbe results on RoboDojo. The unified harness lays the groundwork for generalizable agentic training.
📄 https://modelscope.ai/papers/2611.01560

🏆 Stronger robot use: Qwen3.8-Flash-Next achieves 31.39 Score / 25.00% success without privileged simulator inputs, versus 28.97 / 22.48% for the published GPT-6 Astra–RoboProbe configuration.

🤖 Physical-robot validation: Across 19 tasks, zero-shot Qwen scores 49.50/100, outscoring both π₀.₅ (47.75, fine-tuned on 10 hours of demonstrations per task) and Dream-1 (35.50, evaluated zero-shot). The agent uses depth; both baselines use RGB only.

🛠️ Unified for generalizable agentic training: Shared motion, perception, skill, and episode interfaces make trajectories across environments and robot embodiments easier to combine for future training.

🧠 What to train next: Analysis identifies visual understanding, spatial grounding, geometric coding, and embodied reasoning as joint priorities for future training.

![](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Fmedia%2FHWNadFya8AAE11Q.jpg&mode=thumb&exp=1793923200&sig=9de10e7270ac4a1c)

## Quoted post @zuo_yuxin

> ‼️ Can Qwen directly control the robot as well as GPT-6 Astra?
> 
> 🚀 Introducing uni-robot-agent: a unified harness for Robot-Use Agents (RUAs), a first step toward RUA training and local deployment.
> 
> We tested uni-robot-agent with Qwen3.8-Flash-Next running locally:
> 1️⃣ On RoboDojo, its score rose from 1.90 to 31.39, surpassing the RoboDojo Team’s score with GPT-6 Astra.
> 2️⃣ On real-world tasks, it outscored fine-tuned π0.5 by 5.75 points and zero-shot Dream-1 by 10 points.
> 🧵 [1/7]

https://x.com/zuo_yuxin/status/2117635294115795134