Atlas

Benchmarks

← All benchmarks

RuneBench Skill - Defence

Games · 2026-02-26

RuneBench 30-minute Defence skill task. Agents play RuneScape through the SDK and are scored by the best real-game Defence XP per minute achieved in any 15-second window.

Top models (higher is better)

ModelScore
GPT-6 Astra624
GPT-6.1 Sol590
Gemini 4 Argon460
GPT-6 Sol448
Claude Fable 5438
Claude Opus 5.5332
GPT-5.6 Sol324
Gemini 3.7 Flash308
Grok 4.6270
Claude Fable 5.1258
Gemini 3.8 Flash258
GPT-5.6 Terra238
Claude Sonnet 5.5226
Muse Spark 1.2204
Claude Opus 5194
Gemini 3.6 Flash194
Opus 4.8190
Gemini 3.5 Flash178
Muse Spark 1.1170
GPT-5.5162
Gemini 3 Flash Preview158
Grok 4.7148
Mistral Large 4 Preview142
DeepSeek V4.1 Flash132
GPT-6 Luna116
Qwen3.7-Max (source unspecified)114
GPT-5.6 Luna112
GLM-5.2108
Opus 4.7102
Gemini 3.1 Pro Preview102
Kimi K2.7 Code102
GLM-5.3100
Grok 4.5100
Kimi K2.698.0
Opus 4.694.0
Sonnet 4.692.0
Sonnet 592.0
Haiku 4.586.0
DeepSeek-V4-Flash-073186.0
Grok 4.386.0
Loading Atlas data…