nuta/nsh
A command-line shell like fish, but POSIX compatible.
1,960
Generated Behavioral Tests
83.9%
Best Score
Gemini 3.1 Pro
Results by Model
Hover a point for details · The line marks the Pareto frontier (best score per cost) · Click a point to see model details
Per-Model Results
18 runs| # | Model | Score | Cost | Calls | ||
|---|---|---|---|---|---|---|
| 1 |  |
Gemini 3.1 Pro | 83.9% | $0.92 | 64 | trace → |
| 2 |  |
GLM-5.2 | 83.6% | $38.51 | 225 | trace → |
| 3 |  |
GPT 5.6 Sol (xhigh) | 82.0% | $7.07 | 47 | trace → |
| 4 | |
GPT 5.5 (xhigh) | 81.1% | $9.98 | 81 | trace → |
| 5 |  |
Claude Sonnet 4.6 | 80.6% | $27.07 | 496 | trace → |
| 6 | |
Claude Opus 4.8 (xhigh) | 78.6% | $27.98 | 158 | trace → |
| 7 | |
Claude Opus 4.6 | 76.8% | $13.22 | 304 | trace → |
| 8 | |
GPT 5.5 (high) | 76.3% | $2.79 | 42 | trace → |
| 9 | |
GPT 5 mini | 76.2% | $0.01 | 10 | trace → |
| 10 | |
Claude Opus 4.7 (xhigh) | 71.3% | $5.86 | 116 | trace → |
| 11 | |
Claude Opus 4.7 | 66.2% | $3.70 | 92 | trace → |
| 12 | |
Gemini 3.6 Flash | 55.2% | $7.28 | 204 | trace → |
| 13 | |
GPT 5.5 | 53.0% | $1.33 | 23 | trace → |
| 14 | |
GPT 5.4 | 50.3% | $0.19 | 8 | trace → |
| 15 | |
Gemini 3.5 Flash | 41.7% | $7.17 | 192 | trace → |
| 16 | |
Claude Haiku 4.5 | 36.8% | $1.16 | 148 | trace → |
| 17 | |
Gemini 3 Flash | 31.3% | $0.28 | 122 | trace → |
| 18 | |
GPT 5.4 mini | 0.0% | $0.02 | 14 | trace → |
Click a row to replay how that model rebuilt this program, or the model name to open its full run