Web Navigation on WebArena Lite
61.8Average Success RateQwen3-30B-A3B
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Qwen3-30B-A3BMethod=RL2025.10 | 61.8 | — | 53.3 | 57.1 | — | — | — | 71.1 | — | — | |
| Qwen3-30B-A3BMethod=SFT2025.10 | 58.2 | — | 53.3 | 54.3 | — | — | — | 64.4 | — | — | |
| Qwen3-4BMethod=RL2025.10 | 57.3 | — | 56.7 | 54.3 | — | — | — | 60 | — | — | |
| Qwen3-4BMethod=SFT2025.10 | 54.1 | — | 53.3 | 51.4 | — | — | — | 56.7 | — | — | |
| Llama-3.1-8B + WebAgent-R1Method=RL2025.10 | 51.8 | — | 56.7 | 57.1 | — | — | — | 44.4 | — | — | |
| Claude 4.5 SonnetMethod=Single-Prompt2025.10 | 50 | — | 50 | 62.9 | — | — | — | 40 | — | — | |
| Gemma3 + MiRA#Params=12B2026.03 | 43 | 73.7 | 56.7 | 54.3 | 30.8 | 28.9 | — | — | — | — | |
| Llama3.1 + WebRL#Params=8B2026.03 | 38.8 | 68.4 | 36.7 | 51.4 | 15.4 | 40 | — | — | — | — | |
| SFT collector with SRCDeployment Mode=teacher-assisted data collection2026.06 | 36 | — | — | — | — | — | — | — | 31.24 | 11.9 | |
| Qwen3-30B-A3BMethod=Single-Prompt2025.10 | 35.5 | — | 43.3 | 40 | — | — | — | 26.7 | — | — | |
| Gemma3 + WebRL#Params=12B2026.03 | 35.1 | 68.4 | 43.3 | 40 | 30.8 | 20 | — | — | — | — | |
| Gemma3 + DigiRL#Params=12B2026.03 | 33.3 | 52.6 | 46.7 | 37.1 | 34.6 | 20 | — | — | — | — | |
| Gemini-2.5-pro-SGO#Params=N/A2026.03 | 32.1 | 26.3 | 50 | 37.1 | 23.1 | 28.9 | — | — | — | — | |
| Gemma3 + SFT (BC)#Params=12B2026.03 | 30.9 | 52.6 | 40 | 42.9 | 23.1 | 17.8 | — | — | — | — | |
| Llama3.1 + DigiRL#Params=8B2026.03 | 30.3 | 57.9 | 26.7 | 37.1 | 33.3 | 17.8 | — | — | — | — | |
| Kimi K2.5Deployment Mode=teacher-free test-time policy2026.06 | 29 | — | — | — | — | — | — | — | 32.05 | — | |
| Llama3.1 + AWR#Params=8B2026.03 | 28.5 | 57.9 | 26.7 | 31.4 | 26.7 | 17.8 | — | — | — | — | |
| LEAP-style post-hoc correctionDeployment Mode=teacher-assisted data collection2026.06 | 26.8 | — | — | — | — | — | — | — | 28.64 | 28.64 | |
| Qwen3-4BMethod=Single-Prompt2025.10 | 24.5 | — | 30 | 22.9 | — | — | — | 22.2 | — | — | |
| Expert SFT + SRCDeployment Mode=teacher-free test-time policy2026.06 | 24 | — | — | — | — | — | — | — | 30.13 | — | |
| Gemini-2.5-pro#Params=N/A2026.03 | 23 | 31.6 | 30 | 34.3 | 15.4 | 13.3 | — | — | — | — | |
| GLM-4 + SFT (BC)#Params=9B2026.03 | 22.4 | 47.4 | 13.3 | 31.4 | 23.3 | 13.3 | — | — | — | — | |
| Gemini-2.5-flash#Params=N/A2026.03 | 20.6 | 21.1 | 33.3 | 14.3 | 11.5 | 24.4 | — | — | — | — | |
| Llama3.1 + SFT (BC)#Params=8B2026.03 | 20.6 | 36.8 | 6.7 | 20 | 33.3 | 17.8 | — | — | — | — | |
| Expert SFTDeployment Mode=teacher-free test-time policy2026.06 | 20.5 | — | — | — | — | — | — | — | 15.34 | — | |
| AutoWebGLM#Params=6B2026.03 | 18.2 | 9.4 | 15 | 28.6 | 24.8 | 17.1 | — | — | — | — | |
| GPT-4-Turbo#Params=N/A2026.03 | 17.6 | 10.5 | 16.7 | 14.3 | 36.7 | 13.3 | — | — | — | — | |
| OEC-style random switchDeployment Mode=teacher-assisted data collection2026.06 | 17 | — | — | — | — | — | — | — | 16.61 | 10.15 | |
| Base modelDeployment Mode=teacher-free test-time policy2026.06 | 15.2 | — | — | — | — | — | — | — | 18.48 | — | |
| GPT-4o#Params=N/A2026.03 | 13.9 | 10.5 | 10 | 20 | 20 | 11.1 | — | — | — | — | |
| GLM-4-Chat#Params=9B2026.03 | 6.1 | 5.3 | 10 | 6.7 | 3.3 | 6.7 | — | — | — | — |