Web Task Completion on REAL WebArena Overall
93AccuracyJIT-Planner
Evaluation results
| Method | Links | |
|---|---|---|
| JIT-Planner2026.05 | 93 | |
| Browser-Usecache=true2026.05 | 84 | |
| OpenAI CUA2026.05 | 75 | |
| Anthropic CUA2026.05 | 72 | |
| Browser-Usecache=false2026.05 | 65 |
| Method | Links | |
|---|---|---|
| JIT-Planner2026.05 | 93 | |
| Browser-Usecache=true2026.05 | 84 | |
| OpenAI CUA2026.05 | 75 | |
| Anthropic CUA2026.05 | 72 | |
| Browser-Usecache=false2026.05 | 65 |