NetConfArena - Test-Case Score: leaderboard
Metric: Mean fraction of task test cases satisfied (%). Source: arxiv.org. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V4 Pro (Non-reasoning) | 96.1 |
| 2 | DeepSeek V4 Flash (Non-reasoning) | 91.4 |
| 3 | Qwen 3 32B (Thinking) | 72.8 |
| 4 | Qwen 3 32B (Non-reasoning) | 65.6 |
| 5 | Qwen 3 8B (Thinking) | 45.7 |
| 6 | Qwen 3 8B (Non-reasoning) | 33.7 |
Interactive version: theaggregate.ai/benchmark?slug=netconfarena-test-case-score · How It Works · Data refreshed daily, snapshot 2026-09-19.