Share your thoughts, 1 month free Claude Pro on usSee more

Single-step action prediction on Cloud Console Benchmark 400-sample

95.25Accuracy

Gemini 3 Pro Preview

Updated 1mo ago

Evaluation Results

Method	Links
Gemini 3 Pro Preview 2026.06		95.25
GPT-5.5 2026.06		94.75
AliyunConsoleAgent-32B 2026.06		92.75
Kimi K2.6 2026.06		92.33
Qwen3.6-Plus 2026.06		90.75
AliyunConsoleAgent (Qwen3-VL-8B) 2026.06		88.84
Qwen3-VL-32B-Instruct 2026.06		83
Qwen3-VL-8B-Instruct 2026.06		71.17