Agentic Phone Use on AndroidWorld
83.2Task Success RatePhoneBuddy-4B-Real+Mock
Evaluation Results
| Method | Links | |
|---|---|---|
| PhoneBuddy-4B-Real+MockParameters=4B, Training=mixed RL in real-app and mock-app2026.06 | 83.2 | |
| Gemini 3.1 Pro2026.06 | 80.2 | |
| PhoneBuddy-4B-RealParameters=4B, Training=real-app RL2026.06 | 77.2 | |
| Seed 2.0 Pro2026.06 | 71.5 | |
| GPT-5.42026.06 | 70.7 | |
| PhoneBuddy-4B-SFTParameters=4B, Training=supervised fine-tuning baseline2026.06 | 60.3 | |
| Claude Opus 4.72026.06 | 56 |