Échecs AI BENCHY
Échecs Appel d'outil invalide
Voyez quels modèles d'IA rencontrent le plus souvent Appel d'outil invalide, pour repérer les risques de fiabilité avant de choisir. Trier par: Nombre d'échecs ↑.
| Rang | Modèle | Entreprise | Nombre de Appel d'outil invalide | Score | Tests corrects | Temps de réponse (moy.) |
|---|---|---|---|---|---|---|
| #32 | Gemini 3.5 Flash minimal | 1 | 7.7 | 14/21 | 1.57s | |
| #78 | Qwen3.6 27B medium | Qwen | 1 | 6.8 | 10/21 | 59.7s |
| #106 | Grok 4.20 Beta none | X AI | 1 | 5.8 | 6/18 | 1.19s |
| #107 | Laguna Xs.2 medium | Poolside | 1 | 5.8 | 6/19 | 6.73s |
| #112 | GLM 5.1 none | Z.ai | 1 | 5.7 | 7/21 | 4.10s |
| #118 | Qwen3.6 27B none | Qwen | 1 | 5.6 | 7/21 | 3.72s |
| #119 | Cobuddy medium | Baidu | 1 | 5.6 | 7/21 | 39.9s |
| #122 | GLM 4.7 Flash none | Z.ai | 1 | 5.5 | 6/21 | 2.86s |
| #127 | Grok 4.20 none | X AI | 1 | 5.4 | 6/18 | 1.11s |
| #128 | Qwen3.6 Flash none | Qwen | 1 | 5.4 | 7/21 | 1.60s |
| #129 | MiniMax M2.5 medium | Minimax | 1 | 5.3 | 5/21 | 65.4s |
| #130 | MiniMax M2.7 medium | Minimax | 1 | 5.3 | 5/21 | 38.2s |
| #133 | DeepSeek V3.2 none | DeepSeek | 1 | 5.2 | 6/21 | 13.8s |
| #136 | Elephant Alpha medium | Openrouter | 1 | 5.1 | 6/21 | 1.27s |
| #137 | Elephant Alpha none | Openrouter | 1 | 5.1 | 5/21 | 1.22s |