RuneBench Skill - Fishing
Games · 2026-02-26
RuneBench 30-minute Fishing skill task. Agents play RuneScape through the SDK and are scored by the best real-game Fishing XP per minute achieved in any 15-second window.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 4 Argon | 3587 |
| Claude Opus 5.5 | 1340 |
| Grok 4.6 | 1300 |
| Claude Opus 5 | 1245 |
| GPT-6 Astra | 1220 |
| Gemini 3.8 Flash | 1200 |
| Claude Fable 5 | 1195 |
| Gemini 3.6 Flash | 1145 |
| GLM-5.2 | 1125 |
| Grok 4.5 | 1115 |
| Gemini 3 Flash Preview | 1085 |
| GPT-6.1 Sol | 1080 |
| DeepSeek V4.1 Flash | 1035 |
| GPT-6 Sol | 1010 |
| Claude Fable 5.1 | 995 |
| Grok 4.7 | 930 |
| GLM-5.3 | 890 |
| Muse Spark 1.1 | 890 |
| Gemini 3.5 Flash | 835 |
| GPT-5.4 Mini | 830 |
| GPT-6 Luna | 820 |
| GPT-5.6 Sol | 810 |
| Muse Spark 1.2 | 805 |
| GPT-5.6 Terra | 785 |
| GPT-5.5 | 780 |
| GPT-5.4 | 755 |
| Gemini 3.7 Flash | 685 |
| Opus 4.7 | 585 |
| Claude Sonnet 5.5 | 570 |
| Muse Spark 1.3 | 565 |
| DeepSeek-V4-Flash-0731 | 530 |
| Muse Wonder | 495 |
| Opus 4.5 | 445 |
| Claude Haiku 5.5 | 440 |
| GPT-5.6 Luna | 425 |
| DeepSeek-V4-Pro | 415 |
| Inkling | 390 |
| GLM-5.3 Flash | 370 |
| Kimi K3 | 365 |
| Gemini 3.5 Flash-Lite | 360 |