Software Engineering on Terminal-Bench full 80-task old-core set 1.1
42Resolution RateSpell Opus 4.6
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Spell Opus 4.6Config=default, Effort=medium2026.05 | 42 | 54.98 | 13.38 | |
| Claude Code Opus 4.6Config=default, Effort=medium2026.05 | 42 | 53.96 | 33.26 |