Gert Labs Rankings: leaderboard
Gert Labs global model ranking across game environments that evaluate agentic coding, one-shot coding, and decision-making performance.
Metric: GScore (%). Source: gertlabs.com. Status: years away from saturation. 100 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5.1 | 74.45 |
| 2 | Claude Opus 5 | 73.66 |
| 3 | Claude Fable 5 | 72.25 |
| 4 | GPT-5.6 Sol | 70.95 |
| 5 | Grok 4.6 | 69.35 |
| 6 | Gemini 3.7 Flash | 65.32 |
| 7 | GPT-5.5 | 63.82 |
| 8 | GLM-5.3 | 61.92 |
| 9 | Kimi K3 | 60.91 |
| 10 | GPT-5.6 Terra | 60.91 |
| 11 | Gemini 3.6 Flash | 60.41 |
| 12 | Muse Spark 1.1 | 60.18 |
| 13 | DeepSeek V4 Pro (0813) | 59 |
| 14 | Claude Opus 4.8 | 58.29 |
| 15 | Qwen 3.8 Max | 57.01 |
Interactive version: theaggregate.ai/benchmark?slug=gert-labs-rankings · How It Works · Data refreshed daily, snapshot 2026-09-05.