APEX-Agents: leaderboard
The AI Productivity Index for Agents (APEX-Agents) measures whether frontier AI agents can execute long-horizon, cross-application tasks across three jobs in professional services.
Metric: Mean Score (ReAct) (self-reported). Source: benchmarklist.com. Status: saturation imminent. 45 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.5 Flash | 66.1 |
| 2 | GPT-5.4 (xHigh) | 52.7 |
| 3 | Claude Opus 4.7 (Max) | 50.6 |
| 4 | GPT-5.2 (xHigh) | 48.4 |
| 5 | Claude Opus 4.6 (Max) | 48.4 |
| 6 | Gemini 3.1 Pro (Preview) (High) | 48.2 |
| 7 | GPT-5.3 Codex (High) | 46.9 |
| 8 | Claude Opus 4.6 (High) | 45.6 |
| 9 | Kimi K3 (Max) | 41 |
| 10 | Claude Sonnet 4.6 (High) | 40.7 |
| 11 | GPT-5.6 Sol (Max) | 39.9 |
| 12 | Gemini 3 Flash (High) | 39.5 |
| 13 | Claude Opus 4.8 (Max) | 39.4 |
| 14 | GPT-5.5 (xHigh) | 38.5 |
| 15 | GPT-5.4 Mini (xHigh) | 37.5 |
Interactive version: theaggregate.ai/benchmark?slug=apex-agents · How It Works · Data refreshed daily, snapshot 2026-09-05.