Benchmarks · reasoning
Chartography
Surge AI's 100 professional chart-reading prompts across 12 domains: specialized graphics (Kaplan-Meier curves, candlesticks, contour maps, Sankey diagrams, Bode plots) that need perception, domain knowledge and multi-step reasoning. Accuracy.
42 models · 42 results · data as of 2026-10-29 · surgehq.ai ↗
Leaderboard
Gemini 4 Carbon high
All results
42 rows| # | Model | Effort | Chartography |
|---|---|---|---|
| 1 | Gemini 4 Carbon | high | 72.5%launch |
| 2 | Gemini 4 Argon | high | 71.6% |
| 3 | GPT-6 Astra | max | 71.0% |
| 4 | Opus 5.5 | max | 66.3% |
| 5 | GPT-6 Sol | max | 53.6% |
| 6 | Fable 5.1 | max | 46.2% |
| 7 | GPT-5.6 Sol | max | 45.0% |
| 8 | Gemini 3.8 Flash | high | 40.9% |
| 9 | Gemini 3.7 Flash | high | 40.4% |
| 10 | Fable 5 | max | 34.8% |
| 11 | Gemini 3.6 Flash | medium | 34.0% |
| 12 | GPT-5.6 Terra | max | 34.0% |
| 13 | Muse Spark 1.2 | xhigh | 32.1% |
| 14 | GPT-5.6 Luna | max | 31.4% |
| 15 | GPT-5.5 | xhigh | 31.0% |
| 16 | GPT-5.4 | xhigh | 29.6% |
| 17 | GPT-6 Luna | max | 29.1% |
| 18 | Qwen 3.8 Max | xhigh | 29.1% |
| 19 | Muse Spark 1.3 | xhigh | 27.6% |
| 20 | Opus 5 | max | 27.3% |
| 21 | Kimi K3 | max | 26.6% |
| 22 | Gemini 3.1 Pro | high | 26.1% |
| 23 | Muse Spark 1.1 | xhigh | 24.4% |
| 24 | Grok 4.6 | xhigh | 20.5% |
| 25 | Qwen 3.8 Flash | xhigh | 17.4% |
| 26 | Grok 4.5 | high | 17.0% |
| 27 | Sonnet 5 | max | 16.6% |
| 28 | Opus 4.7 | max | 16.5% |
| 29 | GLM 5.3 Flash | max | 16.3% |
| 30 | Opus 4.8 | max | 15.9% |