Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Software Engineering on Terminal-Bench 2.0
Loading...
75.1
Public Score
GPT-5.4
51.18
57.39
63.6
69.81
Jun 11, 2026
Public Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
Public Score
GPT-5.4
Harness=Proprietary
2026.06
75.1
Claude Opus 4.7
Harness=Terminus-2
2026.06
69.4
Gemini 3.1 Pro
Harness=Terminus-2
2026.06
68.5
GPT-5.2
Harness=Codex CLI
2026.06
62.9
Claude Opus 4.6
Harness=Claude Code
2026.06
58
Qwen3.5-397B-A17B
Harness=Terminus-2
2026.06
52.5
Claude Opus 4.6
Harness=OpenCode
2026.06
52.1
Feedback
Search any
task
Search any
task