LLM Stats (AutomationBench): leaderboard
Metric: Score (%). Source: llm-stats.com. 18 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Muse Spark 1.3 | 49.4 |
| 2 | GLM-5.3 Flash | 48.8 |
| 3 | GLM-5.3 | 48.2 |
| 4 | GPT-6 | 41.4 |
| 5 | Hy4 preview | 32.1 |
| 6 | DeepSeek V4 Pro (0813) | 31.8 |
| 7 | Claude Fable 5.1 | 31.4 |
| 8 | Kimi K3 | 30.8 |
| 9 | Gemini 3.7 Flash | 30.4 |
| 10 | Qwen 3.8 Max | 27.3 |
| 11 | Claude Opus 5 | 26 |
| 12 | DeepSeek-V4-Flash-Vision-Exp | 25.7 |
| 13 | DeepSeek V4 Flash (0731) | 25.1 |
| 14 | GPT-5.6 Sol | 18.1 |
| 15 | Claude Fable 5 | 17.4 |
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-automationbench · How It Works · Data refreshed daily, snapshot 2026-09-05.