README.md

---
license: mit
tags:
- benchmark
- llm-eval
- openrouter
---
# LLM Benchmark — Sultan Command Eval Suite

Automated by JobWeaver / Sultan Command agent orchestration.

## Dry-Run Results (reduced dataset)

| Model | Overall | MMLU-Pro | MATH-500 | BFCL | HumanEval |
|-------|---------|----------|----------|------|-----------|
| DeepSeek V4 Flash | 100.0% | 100.0% | 100.0% | 100.0% | 100.0% |
| Claude Sonnet 4 | 87.5% | 100.0% | 50.0% | 100.0% | 100.0% |

## Charts

![Category Breakdown](benchmark_category_chart.png)
![Score vs Speed](benchmark_score_vs_speed.png)
![Radar](benchmark_radar.png)

## Methodology

- Temperature: 0.2 | Dry-run (reduced question set)
- All models via OpenRouter API
- Full benchmark with 7+ models and standard datasets coming soon

*Generated by Sultan Command on 2026-06-09*