facebook/zstd
Zstandard - Fast real-time compression algorithm
2,038
Generated Behavioral Tests
70.8%
Best Score
Gemini 3.6 Flash
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
19 runs| # | Model | Score | Cost | Calls | ||
|---|---|---|---|---|---|---|
| 1 |  |
Gemini 3.6 Flash | 70.8% | $3.23 | 144 | trace → |
| 2 |  |
Claude Sonnet 4.6 | 68.8% | $15.29 | 375 | trace → |
| 3 |  |
GPT 5.5 (xhigh) | 68.6% | $5.13 | 64 | trace → |
| 4 | |
GPT 5.6 Sol (xhigh) | 68.3% | $7.54 | 53 | trace → |
| 5 | |
GPT 5.5 (high) | 62.4% | $2.54 | 34 | trace → |
| 6 | |
Claude Opus 4.7 (xhigh) | 61.6% | $9.84 | 142 | trace → |
| 7 | |
Claude Opus 4.8 (xhigh) | 60.6% | $25.52 | 143 | trace → |
| 8 | |
GPT 5.5 | 58.4% | $1.24 | 22 | trace → |
| 9 | |
Claude Opus 4.7 | 54.9% | $3.68 | 104 | trace → |
| 10 | |
Gemini 3 Flash | 50.8% | $0.21 | 73 | trace → |
| 11 |  |
GLM-5.2 | 50.4% | $21.42 | 229 | trace → |
| 12 | |
Gemini 3.5 Flash | 48.9% | $4.18 | 210 | trace → |
| 13 | |
Gemini 3.1 Pro | 44.9% | $1.11 | 98 | trace → |
| 14 | |
GPT 5.6 Sol (medium) | 40.7% | $0.32 | 8 | trace → |
| 15 | |
GPT 5.4 | 35.3% | $0.21 | 11 | trace → |
| 16 | |
Claude Haiku 4.5 | 33.9% | $0.84 | 129 | trace → |
| 17 | |
GPT 5.4 mini | 7.7% | $0.02 | 5 | trace → |
| 18 | |
GPT 5 mini | 7.0% | $0.02 | 7 | trace → |
| 19 | |
Claude Opus 4.6 | 0.0% | $1.41 | 189 | trace → |
Click a row to replay how that model rebuilt this program, or the model name to open its full run