APEX-v1-extended
Professional Work · 2025-12-10
Mercor APEX-v1-extended is a held-out professional-work benchmark that extends APEX with expert-evaluated tasks. Scores report percentage task performance.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 4 Argon | 81.1 |
| Muse Spark 1.2 | 75.6 |
| Muse Wonder | 75.1 |
| Muse Spark 1.3 | 73.9 |
| Muse Spark Open | 73.9 |
| Claude Fable 5.1 | 73.8 |
| Muse Spark 1.1 | 73.8 |
| GPT-5.6 Sol Pro | 73.1 |
| GPT-5.6 Terra | 69.5 |
| GPT-6.1 Sol | 69.3 |
| GPT-6 Astra | 69.3 |
| Gemini 3.8 Flash | 68.8 |
| Claude Opus 5 | 68.6 |
| GPT-5.5 | 68.4 |
| Kimi K3 | 68.4 |
| DeepSeek V4.1 Flash 1017 | 68.3 |
| Grok 4.6 | 67.7 |
| Gemini 3.7 Flash | 67.0 |
| GPT-5.6 Luna | 66.6 |
| Claude Fable 5 | 66.0 |
| DeepSeek V4 Pro 0813 | 66.0 |
| Gemini 3.6 Flash | 66.0 |
| GPT-5.3-Codex | 65.9 |
| Opus 4.6 | 65.8 |
| Claude Opus 5.5 | 65.8 |
| GPT-5.4 | 65.6 |
| DeepSeek V4.1 Flash | 65.5 |
| Grok 4.7 | 65.4 |
| MiniMax M3.1 Flash | 65.0 |
| Claude Haiku 5.5 | 64.7 |
| GPT-5.2 | 64.7 |
| GPT-6.1 Terra | 64.2 |
| GPT-6 Sol | 63.6 |
| GLM-5.3 | 63.0 |
| Opus 4.8 | 62.7 |
| Opus 4.7 | 62.2 |
| Gemini 3.5 Flash | 62.2 |
| GPT-6 Luna | 62.0 |
| Gemini 3.1 Pro Preview | 61.8 |
| Grok 4.5 | 61.8 |