综合：无效工具调用
综合
无效工具调用
看看哪些 AI 模型在 综合 上最容易遇到 无效工具调用，更快找出薄弱点。
77/77
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 | 分类得分 | 总成本 | 测试正确 | 响应时间（平均） |
|---|---|---|---|---|---|---|---|
| #8 | Qwen3.7 Max medium | Qwen | 1 | 8.7 | $1.116 | 1/2 | 287.8s |
| #11 | Gemini 3.5 Flash low | 1 | 8.2 | $0.433 | 1/2 | 30.0s | |
| #16 | Muse Spark 1.1 medium | Meta | 1 | 8.3 | $1.357 | 1/2 | 42.6s |
| #17 | Claude Fable 5 medium | Anthropic | 1 | 6.5 | $3.478 | 1/2 | 27.5s |
| #23 | Claude Sonnet 5 medium | Anthropic | 1 | 7.3 | $0.922 | 1/2 | 51.9s |
| #24 | Muse Spark 1.1 low | Meta | 1 | 6.6 | $0.647 | 1/2 | 29.4s |
| #28 | Inkling high | Thinkingmachines | 1 | 7.3 | $1.006 | 1/2 | 63.8s |
| #29 | Step 3.7 Flash medium | Stepfun | 1 | 7.3 | $0.515 | 1/2 | 80.9s |
| #34 | GPT-5.6 Terra high | OpenAI | 1 | 8.7 | $1.055 | 1/2 | 13.7s |
| #36 | Qwen3.7 Plus medium | Qwen | 1 | 8.2 | $0.267 | 1/2 | 190.3s |
| #45 | DeepSeek V4 Flash high | DeepSeek | 1 | 6.4 | $0.042 | 1/2 | 104.1s |
| #51 | Nemotron 3 Ultra medium | NVIDIA | 1 | 6.3 | $0.774 | 1/2 | 218.2s |
| #55 | GPT-5.6 Terra low | OpenAI | 1 | 8.7 | $0.519 | 1/2 | 9.68s |
| #56 | GPT-5.4 Mini medium | OpenAI | 1 | 6.9 | $0.756 | 1/2 | 59.6s |
| #57 | Qwen3.5 Plus 2026-02-15 medium | Qwen | 1 | 6.9 | $0.437 | 1/2 | 175.8s |