List your company

GUI Navigation on MMG2Skill-Bench GUI

77.25Success RateMMG2Skill
12.5129.317546.12562.9325Jun 1, 2026

Evaluation results

MethodLinks
MMG2SkillBackbone=Kimi-K2.6, Prompting Variant=MMG2Skill2026.0677.25
Raw GuideBackbone=Kimi-K2.6, Prompting Variant=Raw Guide2026.0674.75
VanillaBackbone=Kimi-K2.6, Prompting Variant=Vanilla2026.0669.75
MMG2SkillBackbone=Claude-Opus-4.6, Prompting Variant=MMG2Skill2026.0664.75
VanillaBackbone=Claude-Opus-4.6, Prompting Variant=Vanilla2026.0662.22
MMG2SkillBackbone=Claude-Sonnet-4.6, Prompting Variant=MMG2Skill2026.0659.76
Raw GuideBackbone=Claude-Opus-4.6, Prompting Variant=Raw Guide2026.0659.75
MMG2SkillBackbone=GPT-5.5, Prompting Variant=MMG2Skill2026.0657.24
VanillaBackbone=Claude-Sonnet-4.6, Prompting Variant=Vanilla2026.0649.76
Raw GuideBackbone=Claude-Sonnet-4.6, Prompting Variant=Raw Guide2026.0647.5
MMG2SkillBackbone=Qwen3.6-Plus, Prompting Variant=MMG2Skill2026.0639.99
VanillaBackbone=GPT-5.5, Prompting Variant=Vanilla2026.0639.74
Raw GuideBackbone=GPT-5.5, Prompting Variant=Raw Guide2026.0637.24
MMG2SkillBackbone=Gemini-3.1-Pro-Preview, Prompting Variant=MMG2Skill2026.0635
Raw GuideBackbone=Gemini-3.1-Pro-Preview, Prompting Variant=Raw Guide2026.0619.99
VanillaBackbone=Gemini-3.1-Pro-Preview, Prompting Variant=Vanilla2026.0619.97
Raw GuideBackbone=Qwen3.6-Plus, Prompting Variant=Raw Guide2026.0617.5
VanillaBackbone=Qwen3.6-Plus, Prompting Variant=Vanilla2026.0615