Neura Intelligence Index
Back to Leaderboard
TAU-Bench Retail
Tool-Agent-User benchmark for retail customer service scenarios.
agent0-100 percent
Model Rankings (25)
| # | Model | Score | Provider | Source |
|---|---|---|---|---|
| 1 | 86.2 | Anthropic | verified | |
| 2 | 82.4 | Anthropic | verified | |
| 3 | 81.4 | Anthropic | verified | |
| 4 | 81.2 | Anthropic | verified | |
| 5 | 80.5 | Anthropic | verified | |
| 6 | Z GLM-4.5🇨🇳 | 79.7 | Zhipu AI | verified |
| 7 | Z GLM-4.5-Air🇨🇳 | 77.9 | Zhipu AI | verified |
| 8 | Q Qwen3-Coder 480B A35B Instruct🇨🇳 | 77.5 | Alibaba Cloud / Qwen Team | verified |
| 9 | 71.8 | OpenAI | verified | |
| 10 | 70.8 | OpenAI | verified | |
| 11 | Q Qwen3-Next-80B-A3B-Thinking🇨🇳 | 69.6 | Alibaba Cloud / Qwen Team | verified |
| 12 | 69.2 | Anthropic | verified | |
| 13 | 68.4 | OpenAI | verified | |
| 14 | 68 | OpenAI | verified | |
| 15 | Q Qwen3-235B-A22B-Thinking-2507🇨🇳 | 67.8 | Alibaba Cloud / Qwen Team | verified |
| 16 | 67.8 | OpenAI | verified | |
| 17 | 67.8 | MiniMax | verified | |
| 18 | 63.5 | MiniMax | verified | |
| 19 | Q Qwen3-Next-80B-A3B-Instruct🇨🇳 | 60.9 | Alibaba Cloud / Qwen Team | verified |
| 20 | 60.3 | OpenAI | verified | |
| 21 | 57.6 | OpenAI | verified | |
| 22 | 55.8 | OpenAI | verified | |
| 23 | 54.8 | OpenAI | verified | |
| 24 | 51 | Anthropic | verified | |
| 25 | 22.6 | OpenAI | verified |