LLM Stats (Tau3 Banking): leaderboard
Banking domain of Tau3-Bench: multi-turn customer-support conversations where the agent must use banking tools correctly and stay inside policy to resolve the case.
Metric: Score (%). Source: llm-stats.com. Saturation forecast: Around October 2027. 13 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Ling-3.0-flash-fin | 41 | #108 |
| 2 | Grok 4.5 | 33 | #48 |
| 3 | Ling-3.0-flash | 28 | #206 |
| 4 | Inkling | 23.7 | #126 |
| 5 | Muse Glimmer 30B | 23.5 | #178 |
| 6 | Solar Pro 4 | 23 | #215 |
| 7 | Nemotron 3.5 Super VL (BF16) | 18.88 | #165 |
| 8 | Inkling Small | 15.5 | #138 |
| 9 | Mistral Medium 3.5 | 13.4 | #353 |
| 10 | Nemotron 3.5 Lightning | 9.28 | #467 |
| 11 | LFM2.5-2.6B | 5.67 | #822 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-tau3-banking · How It Works · Data refreshed daily, snapshot 2026-10-20.