无效工具调用 失败
看看哪些 AI 模型最常遇到 无效工具调用，让你在选择前先发现稳定性风险。 排序方式: 测试正确 ↑.
190/190
筛选模型
没有模型匹配当前搜索和筛选条件。
| 排名 | 模型 | 公司 | 无效工具调用 次数 | 分数 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #173 | Step 3.7 Flash low | Stepfun | 2 | 0.7 | $0.449 | 11/23 | 20.7s |
| #174 | Qwen3.6 27B medium | Qwen | 2 | 0.7 | $1.021 | 11/23 | 107.8s |
| #189 | Muse Glimmer 30B low | Meta | 2 | 0.7 | $0.219 | 11/23 | 22.4s |
| #193 | GPT-5.6 Luna low | OpenAI | 1 | 0.7 | $0.049 | 11/23 | 6.96s |
| #198 | Inkling low | Thinkingmachines | 2 | 0.7 | $0.293 | 11/23 | 6.81s |
| #212 | Apodex 1.1 Mini high | Apodex | 1 | 0.7 | $0.000 | 11/23 | 31.2s |
| #213 | GLM 5.3 Flash low | Z.ai | 1 | 0.7 | $0.038 | 11/23 | 11.8s |
| #219 | Step 3.7 Flash high | Stepfun | 1 | 0.7 | $1.478 | 11/23 | 72.7s |
| #229 | KAT-Coder-Pro V2.5 high | Kwaipilot | 1 | 0.6 | $0.506 | 11/23 | 21.2s |
| #238 | Beam 1.0 501B A23B low | Reflection | 1 | 0.6 | $0.432 | 11/23 | 4.89s |
| #241 | MiMo-V2.5-Pro medium | Xiaomi | 1 | 0.6 | $0.332 | 11/23 | 56.7s |
| #245 | Qwen3.5-35B-A3B medium | Qwen | 1 | 0.6 | $0.716 | 11/23 | 111.5s |
| #251 | Qwen3.5 Plus 2026-02-15 none | Qwen | 1 | 0.6 | $0.176 | 11/23 | 20.1s |
| #255 | DeepSeek V3.2 medium | DeepSeek | 1 | 0.6 | $0.152 | 11/23 | 70.1s |
| #258 | Grok 4.20 medium | X AI | 1 | 0.6 | $1.246 | 11/23 | 36.7s |