Classement Tâches agentiques
Voyez quels modèles d'IA réussissent le mieux sur Tâches agentiques, lesquels restent fiables et où les écarts sont les plus marqués. Trier par: Tests corrects ↓.
380/380
Filtrer les modèles
Aucun modèle ne correspond à la recherche et aux filtres actuels.
| Rang | Modèle | Entreprise | Score Tâches agentiques | Score | Coût total | Tests corrects | Temps de réponse (moy.) |
|---|---|---|---|---|---|---|---|
| #129 | DeepSeek V4 Pro 0423 none | DeepSeek | 10.0 | 7.4 | $0.311 | 1/1 | 65.5s |
| #131 | Gemma 4 26B A4B medium | 10.0 | 7.3 | $0.086 | 1/1 | 107.1s | |
| #136 | LongCat 2.0 high | Meituan | 10.0 | 7.3 | $0.569 | 1/1 | 116.1s |
| #139 | Qwen3.5-27B none | Qwen | 10.0 | 7.2 | $0.116 | 1/1 | 124.5s |
| #148 | Qwen3.6 27B medium | Qwen | 10.0 | 7.2 | $1.006 | 1/1 | 145.9s |
| #150 | GLM 5.3 FlashX medium | Z.ai | 10.0 | 7.1 | $0.114 | 1/1 | 74.4s |
| #152 | MiMo-V2.5 medium | Xiaomi | 10.0 | 7.1 | $0.419 | 1/1 | 113.6s |
| #164 | Gemma 4 31B medium | 10.0 | 6.9 | $0.117 | 1/1 | 82.1s | |
| #165 | GPT-5.6 Luna low | OpenAI | 10.0 | 6.9 | $0.075 | 1/1 | 45.7s |
| #166 | Claude Sonnet 5.5 low | Anthropic | 10.0 | 6.9 | $0.909 | 1/1 | 46.6s |
| #169 | Qwen3.5-Flash medium | Qwen | 10.0 | 6.8 | $0.184 | 1/1 | 67.7s |
| #170 | Inkling low | Thinkingmachines | 10.0 | 6.8 | $0.304 | 1/1 | 44.2s |
| #177 | Qwen3.5-Flash none | Qwen | 10.0 | 6.7 | $0.089 | 1/1 | 53.8s |
| #181 | Ember-1 max | Fireworks | 10.0 | 6.7 | $3.264 | 1/1 | 62.5s |
| #184 | GLM 5.3 Flash low | Z.ai | 10.0 | 6.6 | $0.052 | 1/1 | 59.7s |