| code-generation-on-terminal-bench-3-0 | Claude Opus 5 (Max) (mini-SWE-agent) | #1 | Accuracy: 42.7 |
| code-generation-on-terminal-bench-3-0 | GPT-5.6 Sol (Max) (Codex) | #2 | Accuracy: 34.59 |
| code-generation-on-terminal-bench-3-0 | Claude Fable 5 (Max) (Claude Code) | #3 | Accuracy: 34.05 |
| code-generation-on-terminal-bench-3-0 | GPT-5.6 Terra (Max) (Codex) | #6 | Accuracy: 20.81 |
| code-generation-on-terminal-bench-3-0 | GPT-5.6 Luna (Max) (Codex) | #8 | Accuracy: 14.32 |
| code-generation-on-terminal-bench-3-0 | GLM-5.2 (Max) (Claude Code) | #9 | Accuracy: 4.59 |
| code-generation-on-terminal-bench-3-0 | Claude Opus 4.8 (Max) (Claude Code) | – | Accuracy: 21.08 |
| code-generation-on-terminal-bench-3-0 | Claude Sonnet 5 (Max) (Claude Code) | – | Accuracy: 14.59 |
| code-generation-on-terminal-bench-3-0 | Grok 4.5 (XHigh) (Cursor CLI) | – | Accuracy: 15.68 |
| code-generation-on-terminal-bench-3-0 | Grok 4.6 (High) (Grok Build) | – | Accuracy: 26.49 |
| code-generation-on-terminal-bench-2-0-terminus-2 | GLM-4.7 (Terminus-2) | #9 | Accuracy: 33.4 |
| code-generation-on-terminal-bench-2-0-terminus-2 | GLM-4.6 (Terminus-2) | #11 | Accuracy: 24.5 |