← Back to leaderboard  ·  Show all task instances

eliukblau/pixterm

Draw images in your ANSI terminal with true color

1,014 go easy
430
Generated Behavioral Tests
98.8%
Best Score
Claude Opus 5 (xhigh)

Results by Model

Hover for details · Click a model to open it · Drag to zoom the x-axis · Shift-drag to pan · The line marks the Pareto frontier (best score per cost)

Per-Model Results

25 runs
# Model Score Cost Calls
1 Claude Opus 5 (xhigh) 98.8% $25.36 147 trace →
2 GPT-5.6 Sol (xhigh) 90.2% $10.17 66 trace →
3 GPT 5.5 (xhigh) 85.8% $7.86 63 trace →
4 Gemini 3.5 Flash 80.0% $5.41 114 trace →
5 GLM-5.2 76.7% $46.49 221 trace →
6 Claude Sonnet 4.6 74.9% $24.61 335 trace →
7 GPT 5.5 (high) 74.2% $4.15 42 trace →
8 Claude Opus 4.8 (xhigh) 73.5% $30.02 136 trace →
9 Muse Spark 1.2 (xhigh) 73.5% $0.21 169 trace →
10 Gemini 3.7 Flash 32.3% $2.17 133 trace →
11 GPT-5.6 Sol 32.1% $1.31 14 trace →
12 GPT 5.5 24.7% $1.15 16 trace →
13 Muse Spark 1.3 (xhigh) 21.6% $2.45 304 trace →
14 Gemini 3.1 Pro 21.6% $2.32 78 trace →
15 Claude Opus 4.7 (xhigh) 20.7% $30.03 224 trace →
16 Muse Spark 1.1 (xhigh) 20.7% $1.03 202 trace →
17 Claude Opus 4.6 19.5% $15.01 227 trace →
18 GPT 5.4 19.5% $0.20 7 trace →
19 Gemini 3.6 Flash 19.3% $4.14 136 trace →
20 Claude Opus 4.7 19.3% $5.80 92 trace →
21 GPT 5 mini 17.9% $0.02 14 trace →
22 GPT 5.4 mini 14.4% $0.03 6 trace →
23 Claude Haiku 4.5 9.8% $0.98 136 trace →
24 Gemini 3 Flash 8.4% $0.40 83 trace →
25 Muse Spark 1.3 (max) 0.0% $11.73 358 trace →

Click a row to replay how that model rebuilt this program, or the model name to open its full run