RuneBench Skill - Attack
Games · 2026-02-26
RuneBench 30-minute Attack skill task. Agents play RuneScape through the SDK and are scored by the best real-game Attack XP per minute achieved in any 15-second window.
Top models (higher is better)
| Model | Score |
|---|---|
| GPT-6.1 Sol | 6875 |
| GPT-6 Astra | 2420 |
| Gemini 4 Argon | 2414 |
| Grok 4.7 | 2414 |
| Claude Fable 5.1 | 2412 |
| Claude Opus 5.5 | 446 |
| Gemini 3.5 Flash | 436 |
| GPT-6 Sol | 420 |
| Claude Opus 5 | 352 |
| GPT-5.6 Sol | 350 |
| Grok 4.5 | 326 |
| Gemini 3.8 Flash | 284 |
| Claude Fable 5 | 282 |
| Gemini 3.7 Flash | 254 |
| GPT-5.6 Terra | 246 |
| Grok 4.6 | 210 |
| Sonnet 5 | 196 |
| Opus 4.8 | 192 |
| Claude Sonnet 5.5 | 192 |
| DeepSeek V4.1 Flash | 192 |
| Gemini 3.6 Flash | 168 |
| GPT-5.6 Luna | 126 |
| Gemini 3.5 Flash-Lite | 120 |
| DeepSeek-V4-Flash-0731 | 114 |
| GPT-5.5 | 110 |
| DeepSeek-V4-Pro | 108 |
| Muse Spark 1.1 | 106 |
| Opus 4.7 | 100 |
| Gemini 3 Flash Preview | 96.0 |
| Muse Spark 1.2 | 92.0 |
| Sonnet 4.5 | 86.0 |
| GLM-5.2 | 86.0 |
| Muse Spark 1.3 | 86.0 |
| Inkling | 84.0 |
| Kimi K2.5 | 84.0 |
| Qwen3.8 27B | 84.0 |
| Grok 4.3 | 82.0 |
| GLM-5.3 | 80.0 |
| Qwen3.7-Max (source unspecified) | 74.0 |
| GLM-5.3 Flash | 72.0 |