Agentic Skills Evaluation Framework — leaderboard
Metric: Overall Score w/ (self-reported). Source: benchmarklist.com. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.8 | 92.7 |
| 2 | Claude Opus 4.7 | 92.3 |
| 3 | Claude Sonnet 4.6 | 91.5 |
| 4 | GLM-5.1 | 91.1 |
| 5 | Gemini 3.5 Flash | 88.6 |
| 6 | Gemini 3.1 Pro (Preview) | 88.5 |
| 7 | GPT-5.4 | 88.2 |
| 8 | DeepSeek V4 Pro | 86.6 |
| 9 | Gemini 3 Flash (Preview) | 85.9 |
| 10 | GPT-5.4 Mini | 84.5 |
| 11 | Claude Haiku 4.5 | 84.1 |
| 12 | DeepSeek V4 Flash | 83.9 |
| 13 | GPT-5.4 Nano | 81.9 |
| 14 | Gemini 3.1 Flash Lite | 71.4 |
| 15 | Qwen3 Coder Next | 70.8 |
Interactive version: theaggregate.ai/benchmark?slug=agentic-skills-evaluation-framework · How the rankings work · Data refreshed daily, snapshot 2026-07-22.