APEX-Agents 1.1
Professional Work · 2026-09-08
Mercor APEX-Agents 1.1: 240 tasks across corporate law, management consulting and investment banking, with revised worlds, clearer prompts and a reworked rubric judge. Mean Score is the average fraction of criteria passed. Uses the Loop truncated-tools agent and remains separate from the original 480-task APEX-Agents release.
Top models (higher is better)
| Model | Score |
|---|---|
| Gemini 4 Argon | 87.4 |
| Claude Sonnet 5.5 | 83.7 |
| Claude Opus 5.5 | 81.3 |
| DeepSeek V4.1 Flash 1017 | 79.4 |
| Gemini 3.7 Flash | 79.4 |
| GPT-6.1 Terra | 79.3 |
| Grok 4.6 | 77.3 |
| Claude Fable 5.1 | 77.1 |
| Claude Opus 5 | 77.0 |
| Muse Wonder | 77.0 |
| GPT-6 Astra | 75.1 |
| Claude Fable 5 | 74.5 |
| GPT-6.1 Sol | 73.2 |
| Gemini 3.8 Flash | 72.9 |
| GPT-5.6 Terra | 71.8 |
| Grok 4.5 | 70.5 |
| Muse Spark 1.3 | 70.4 |
| GPT-5.5 | 70.3 |
| GPT-6 Sol | 68.7 |
| Grok 4.7 | 68.7 |
| Claude Haiku 5.5 | 68.3 |
| GPT-5.4 | 68.2 |
| MiniMax M3.1 Flash | 68.1 |
| Sonnet 5 | 67.4 |
| GLM-5.3 Flash | 66.8 |
| GLM-5.3 | 66.2 |
| MiMo-V2.6-Pro | 65.3 |
| GPT-5.6 Sol Pro | 65.2 |
| Opus 4.8 | 64.4 |
| Mistral Large 4 Preview | 64.3 |
| Gemini 3.6 Flash | 63.4 |
| Opus 4.7 | 63.1 |
| Mercury 3 | 62.9 |
| Beam 1.0 501B A23B | 62.5 |
| DeepSeek V4 Pro 0813 | 62.3 |
| Kimi K3 | 62.2 |
| Opus 4.6 | 61.8 |
| GLM-5.2 | 61.1 |
| Muse Spark 1.2 | 59.8 |
| GPT-6 Luna | 59.0 |