List your company

UI Agent Interaction on AgentNetBench

80.2Average SRQwen-3-VL-72B
21.12836.46451.867.136May 11, 2026May 16, 2026May 21, 2026May 27, 2026Jun 1, 2026Jun 6, 2026Jun 12, 2026

Evaluation results

MethodLinks
Qwen-3-VL-72BHistory=5, Drop=✗2026.0580.283.668.351.9—
Qwen-3-VL-72BHistory=3, Drop=✗2026.0579.382.567.450.8—
Qwen3-VL-30B-A3BHistory=5, Drop=✗2026.0577.881.36649.1—
Qwen3-VL-30B-A3BHistory=3, Drop=✗2026.057780.465.148.2—
Qwen-3-VL-72BHistory=1, Drop=✗2026.0576.979.264.548.1—
Qwen3-VL-30B-A3BHistory=1, Drop=✗2026.0574.176.561.344.6—
ReVisionHistory=5, Drop=✓2026.0573.8776144.2—
ReVisionHistory=5, Drop=✗2026.0572.576.259.842.9—
OpenCUAHistory=5, Drop=✗2026.0572.476.159.742.8—
Qwen-3-VL-8BHistory=5, Drop=✗2026.0572.275.259.343.4—
OpenCUAHistory=3, Drop=✗2026.0572.175.859.442.5—
ReVisionHistory=3, Drop=✗2026.057275.959.342.4—
Qwen-3-VL-8BHistory=3, Drop=✗2026.0571.974.658.843.1—
ReVisionHistory=3, Drop=✓2026.0570.774.458.141.7—
UI-TARS-1.5-7BHistory=5, Drop=✗2026.0568.972.354.540.6—
Qwen-3-VL-8BHistory=1, Drop=✗2026.0568.773.454.940.2—
OpenCUAHistory=1, Drop=✗2026.0568.47255.639.8—
UI-TARS-1.5-7BHistory=3, Drop=✗2026.0568.471.853.840.2—
UI-TARS-1.5-7BHistory=1, Drop=✗2026.0566.970.452.338.8—
UI-TARS-72B-DPOHistory=5, Drop=✗2026.0563.867.849.636—
UI-TARS-72B-DPOHistory=3, Drop=✗2026.0563.267.14935.5—
UI-TARS-72B-DPOHistory=1, Drop=✗2026.0562.366.248.134.7—
Kimi-VL-A3BHistory=5, Drop=✗2026.0550.955.240.428.1—
Kimi-VL-A3BHistory=3, Drop=✗2026.0549.85439.527.2—
Kimi-VL-A3BHistory=1, Drop=✗2026.0547.251.637.825.9—
Qwen2.5-VL-72BHistory=5, Drop=✗2026.0541.544.432.421.7—
Qwen2.5-VL-72BHistory=3, Drop=✗2026.0540.843.631.821—
Qwen2.5-VL-32BHistory=5, Drop=✗2026.0538.74130.520.2—
Qwen2.5-VL-72BHistory=1, Drop=✗2026.0538.742.130.520.2—
Qwen2.5-VL-32BHistory=3, Drop=✗2026.0537.940.229.819.5—
Qwen2.5-VL-32BHistory=1, Drop=✗2026.0535.638.528.218.4—
AGMem (Ours)Backbone Model=GPT-5.4-mini2026.0628.8———34.6
+ visual memoryBackbone Model=GPT-5.4-mini2026.0628.1———33.8
Qwen2.5-VL-7BHistory=5, Drop=✗2026.0525.82720.113.3—
Vanilla agentBackbone Model=GPT-5.4-mini2026.0625.8———24.2
Qwen2.5-VL-7BHistory=3, Drop=✗2026.052526.419.512.7—
Qwen2.5-VL-7BHistory=1, Drop=✗2026.0523.425.118.311.9—