Grounding on ScreenSpot V2
94.73Grounding AccuracyGUIrilla-See (7B)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GUIrilla-See (7B)Platform=macOS, Data=Synthetic, # Images=4.2K2025.10 | 94.73 | — | — | |
| Fara1.5-27BModel=Fara1.5-27B2026.06 | 94.4 | — | — | |
| Fara1.5-9BModel=Fara1.5-9B2026.06 | 94 | — | — | |
| Fara1.5-4BModel=Fara1.5-4B2026.06 | 92.3 | — | — | |
| UI-TARS (7B)Platform=Multi-OS, Data=Real + Synthetic, # Images=∼20M (est.)2025.10 | 91.6 | — | — | |
| Full CacheModel=OpenCUA-7B, KV-cache budget=Full Cache2026.06 | 91.13 | — | 51.8 | |
| STaR-KVModel=OpenCUA-7B, KV-cache budget=40%2026.06 | 91.05 | — | 50.04 | |
| PyramidKVModel=OpenCUA-7B, KV-cache budget=40%2026.06 | 91.02 | — | 49.47 | |
| SnapKVModel=OpenCUA-7B, KV-cache budget=40%2026.06 | 91.02 | — | 49.32 | |
| UI-TARS (72B)Platform=Multi-OS, Data=Real + Synthetic, # Images=∼20M (est.)2025.10 | 90.3 | — | — | |
| STaR-KVModel=OpenCUA-7B, KV-cache budget=20%2026.06 | 89.92 | — | 47.01 | |
| GUIrilla-See (3B)Platform=macOS, Data=Synthetic, # Images=4.2K2025.10 | 89.54 | — | — | |
| GUIKVModel=OpenCUA-7B, KV-cache budget=40%2026.06 | 89.1 | — | 47.31 | |
| STaR-KVModel=UI-TARS-1.5-7B, KV-cache budget=40%2026.06 | 88.95 | — | 49.94 | |
| STaR-KVModel=UI-TARS-1.5-7B, KV-cache budget=20%2026.06 | 88.9 | — | 47.31 | |
| PyramidKVModel=OpenCUA-7B, KV-cache budget=20%2026.06 | 88.8 | — | 45.3 | |
| Full CacheModel=UI-TARS-1.5-7B, KV-cache budget=Full Cache2026.06 | 88.79 | — | 49.75 | |
| GUIKVModel=UI-TARS-1.5-7B, KV-cache budget=40%2026.06 | 88.51 | — | 48.92 | |
| GUIKVModel=UI-TARS-1.5-7B, KV-cache budget=20%2026.06 | 88.46 | — | 46.59 | |
| SnapKVModel=UI-TARS-1.5-7B, KV-cache budget=40%2026.06 | 88.34 | — | 47.22 | |
| SnapKVModel=UI-TARS-1.5-7B, KV-cache budget=20%2026.06 | 88.28 | — | 45.09 | |
| GUIKVModel=OpenCUA-7B, KV-cache budget=20%2026.06 | 88.04 | — | 43.03 | |
| PyramidKVModel=UI-TARS-1.5-7B, KV-cache budget=40%2026.06 | 87.58 | — | 45.45 | |
| Qwen 2.5 VL (7B)Platform=Multi-OS, Data=Real + Synthetic, # Images=-2025.10 | 87.5 | — | — | |
| STaR-KVModel=OpenCUA-7B, KV-cache budget=10%2026.06 | 87.01 | — | 42.67 | |
| PyramidKVModel=OpenCUA-7B, KV-cache budget=10%2026.06 | 86.29 | — | 39.65 | |
| GUIKVModel=OpenCUA-7B, KV-cache budget=10%2026.06 | 86.02 | — | 39.16 | |
| STaR-KVModel=UI-TARS-1.5-7B, KV-cache budget=10%2026.06 | 85.39 | — | 40.72 | |
| GUIKVModel=UI-TARS-1.5-7B, KV-cache budget=10%2026.06 | 85.35 | — | 40.7 | |
| SnapKVModel=UI-TARS-1.5-7B, KV-cache budget=10%2026.06 | 85.32 | — | 40.04 | |
| PyramidKVModel=UI-TARS-1.5-7B, KV-cache budget=20%2026.06 | 84.74 | — | 39.46 | |
| UI-TARS (2B)Platform=Multi-OS, Data=Real + Synthetic, # Images=∼20M (est.)2025.10 | 84.7 | — | — | |
| STaR-KVModel=OpenCUA-7B, KV-cache budget=5%2026.06 | 83.58 | — | 35.89 | |
| OS-Atlas (7B)Platform=Multi-OS, Data=Synthetic, # Images=2.2M2025.10 | 83.3 | — | — | |
| GUIKVModel=OpenCUA-7B, KV-cache budget=5%2026.06 | 81.84 | — | 33.78 | |
| PyramidKVModel=OpenCUA-7B, KV-cache budget=5%2026.06 | 81.16 | — | 32.72 | |
| STaR-KVModel=UI-TARS-1.5-7B, KV-cache budget=5%2026.06 | 77.54 | — | 32.74 | |
| ShowUI (2B)Platform=Multi-OS, Data=Real + Synthetic, # Images=256K2025.10 | 77.3 | — | — | |
| GUIKVModel=UI-TARS-1.5-7B, KV-cache budget=5%2026.06 | 77.15 | — | 32.63 | |
| SnapKVModel=UI-TARS-1.5-7B, KV-cache budget=5%2026.06 | 76.71 | — | 32.04 | |
| UGround (7B)Platform=Web + Android, Data=Real, # Images=1.3M2025.10 | 76.3 | — | — | |
| PyramidKVModel=UI-TARS-1.5-7B, KV-cache budget=10%2026.06 | 75.77 | — | 33.12 | |
| SnapKVModel=OpenCUA-7B, KV-cache budget=5%2026.06 | 73.11 | — | 33.62 | |
| SnapKVModel=OpenCUA-7B, KV-cache budget=10%2026.06 | 72.11 | — | 38.27 | |
| SnapKVModel=OpenCUA-7B, KV-cache budget=20%2026.06 | 67.96 | — | 41.17 | |
| Qwen 2.5 VL (3B)Platform=Multi-OS, Data=Real + Synthetic, # Images=-2025.10 | 62.34 | — | — | |
| PyramidKVModel=UI-TARS-1.5-7B, KV-cache budget=5%2026.06 | 57.39 | — | 24.7 | |
| GUIrilla-See (0.7B)Platform=macOS, Data=Synthetic, # Images=4.2K2025.10 | 53.55 | — | — | |
| CogAgent (18B)Platform=Multi-OS, Data=Real + Synthetic, # Images=40M2025.10 | 52.8 | — | — | |
| Florence-2 (0.7B)Platform=Multi-OS, Data=Real + Synthetic, # Images=-2025.10 | 1.8 | — | — | |
| AguvisSize=8B, TTFT ↓ (ms)=119.7, TPOT ↓ (ms/token)=21.22026.04 | — | 85.6 | — | |
| ClaudeVersion=3.72026.04 | — | 87.6 | — | |
| CogAgentSize=18B, TTFT ↓ (ms)=1253.2, TPOT ↓ (ms/token)=208.82026.04 | — | 49.2 | — | |
| FaraParameters=7B2026.04 | — | 89.3 | — | |
| Ferret-UISize=8B, TTFT ↓ (ms)=152.5, TPOT ↓ (ms/token)=22.92026.04 | — | 7.8 | — | |
| Ferret-UISize=3B, TTFT ↓ (ms)=69.5, TPOT ↓ (ms/token)=9.82026.04 | — | 1.9 | — | |
| GeminiVersion=3-Pro2026.04 | — | 93.7 | — | |
| GoClick-BSize=0.2B, TTFT ↓ (ms)=37.7, TPOT ↓ (ms/token)=4.12026.04 | — | 75.2 | — | |
| GoClick-LSize=0.8B, TTFT ↓ (ms)=91.1, TPOT ↓ (ms/token)=8.32026.04 | — | 81.1 | — | |
| GPT-4o2026.04 | — | 20.4 | — | |
| GTA1-32Bdesign=end-to-end2026.02 | — | 95.2 | — | |
| GTA1-7Bdesign=end-to-end2026.02 | — | 92.4 | — | |
| GUI-Owl-32Bdesign=end-to-end2026.02 | — | 93.1 | — | |
| GUI-Owl-7Bdesign=end-to-end2026.02 | — | 92.8 | — | |
| Holo1Parameters=7B2026.04 | — | 89.9 | — | |
| Holo2-30B-A3Bdesign=end-to-end2026.02 | — | 94.9 | — | |
| Holo2-8Bdesign=end-to-end2026.02 | — | 93.2 | — | |
| MAI-UI-2Bdesign=end-to-end2026.02 | — | 92.5 | — | |
| MAI-UI-32Bdesign=end-to-end2026.02 | — | 96.5 | — | |
| MAI-UI-8Bdesign=end-to-end2026.02 | — | 95.2 | — | |
| MolmoWebParameters=4B2026.04 | — | 89.5 | — | |
| MolmoWeb-GroundParameters=8B2026.04 | — | 91.8 | — | |
| OpenAI CUA2026.04 | — | 87.9 | — | |
| OpenCUA-32Bdesign=end-to-end2026.02 | — | 93.4 | — | |
| OpenCUA-72Bdesign=end-to-end2026.02 | — | 92.9 | — | |
| OpenCUA-7Bdesign=end-to-end2026.02 | — | 92.3 | — | |
| OS-ATLAS-4BSize=4B, TTFT ↓ (ms)=137.3, TPOT ↓ (ms/token)=31.42026.04 | — | 68.7 | — | |
| OS-ATLAS-8BSize=8B, TTFT ↓ (ms)=137.5, TPOT ↓ (ms/token)=19.92026.04 | — | 84.1 | — | |
| Qwen2-VLSize=2B, TTFT ↓ (ms)=58.8, TPOT ↓ (ms/token)=16.42026.04 | — | 18.6 | — | |
| Qwen2.5VLParameters=7B2026.04 | — | 89.3 | — | |
| Qwen2VL-7BSize=8B, TTFT ↓ (ms)=118.9, TPOT ↓ (ms/token)=21.22026.04 | — | 66.9 | — | |
| Qwen3-VL-2Breproduction=true, design=end-to-end2026.02 | — | 85.6 | — | |
| Qwen3-VL-30B-A3Breproduction=true, design=end-to-end2026.02 | — | 91.7 | — | |
| Qwen3-VL-8Breproduction=true, design=end-to-end2026.02 | — | 92.1 | — | |
| SeeClickSize=10B, TTFT ↓ (ms)=160.4, TPOT ↓ (ms/token)=184.42026.04 | — | 54 | — | |
| ShowUISize=2B, TTFT ↓ (ms)=79.7, TPOT ↓ (ms/token)=14.72026.04 | — | 77.4 | — | |
| Step-GUI-4Bdesign=end-to-end2026.02 | — | 93.6 | — | |
| Step-GUI-8Bdesign=end-to-end2026.02 | — | 95.1 | — | |
| UGroundParameters=7B2026.04 | — | 76.3 | — | |
| UGroundSize=7B, TTFT ↓ (ms)=1034.6, TPOT ↓ (ms/token)=27.92026.04 | — | 76.5 | — | |
| UI-TARS-1.5-7Bdesign=end-to-end2026.02 | — | 91.6 | — | |
| UI-Venus-1.0-72Bdesign=end-to-end2026.02 | — | 95.3 | — | |
| UI-Venus-1.0-7Bdesign=end-to-end2026.02 | — | 94.1 | — | |
| UI-Venus-1.5-2Bdesign=end-to-end2026.02 | — | 92.8 | — | |
| UI-Venus-1.5-30B-A3Bdesign=end-to-end2026.02 | — | 96.2 | — | |
| UI-Venus-1.5-8Bdesign=end-to-end2026.02 | — | 95.9 | — |