TSCG 50 Tools: leaderboard
Metric: Scenario D, 50 tools, TSCG compilation: tool-selection accuracy (%) (self-reported). Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemma 3 12B | 98 |
| 2 | phi-4-14B | 90.3 |
| 3 | Qwen 3 14B | 89.6 |
| 4 | Gemma 3 4B | 87.4 |
| 5 | Llama 3.1 8B | 86.3 |
| 6 | Qwen 3 4B | 75 |
| 7 | Mistral 7B | 65 |
Interactive version: theaggregate.ai/benchmark?slug=tscg-50-tools · How It Works · Data refreshed daily, snapshot 2026-10-09.