gemini-4IMAGE

Gemini 4 Argon scores 81.9% on RoboVista robotics understanding eval

An evaluation reports Gemini 4 Argon reaches 81.9% on RoboVista robotics understanding, beating GPT-6 Astra and Claude Fable 5.1 at the lowest eval cost.

Shuangyu Xie

@Shuangyu0_0XieArchived 22 days agoOriginally posted 2026.10.16

Gemini 4 Argon scores 81.9% on RoboVista robotics understanding eval (image 1 of 2)

Gemini 4 Argon scores 81.9% on RoboVista robotics understanding eval (image 2 of 2)

Shuangyu Xie @Shuangyu0_0XieOriginally posted 2026.10.16

Open original on X

Gemini-4 Argon shows a large improvement in robotics understanding on RoboVista. It reaches 81.9%, beats GPT-6 Astra and Claude Fable 5.1 in our evaluation with lowest eval cost. Compared with Gemini 3 Flash, it improves across every robotics domain. Example VQA👇

This post quotes another post: x.com/shuangyu0_0xie/status/2076737025437528268

x.com/shuangyu0_0xie/status/2110772438619308517Archived 2026.10.16 10:14