Classement Programmation
Voyez quels modèles d'IA réussissent le mieux sur Programmation, lesquels restent fiables et où les écarts sont les plus marqués. Trier par: Tests corrects ↓.
289/289
Filtrer les modèles
Aucun modèle ne correspond à la recherche et aux filtres actuels.
| Rang | Modèle | Entreprise | Score Programmation | Score | Coût total | Tests corrects | Temps de réponse (moy.) |
|---|---|---|---|---|---|---|---|
| #53 | DeepSeek V4 Flash 0731 low | DeepSeek | 7.3 | 8.0 | $0.072 | 1/3 | 163.3s |
| #54 | DeepSeek V4 Flash 0731 high | DeepSeek | 6.3 | 8.0 | $0.134 | 1/3 | 252.7s |
| #55 | GLM 5.2 high | Z.ai | 6.4 | 8.0 | $1.188 | 1/3 | 73.0s |
| #62 | Qwen3.7 Plus medium | Qwen | 6.1 | 7.9 | $0.277 | 1/3 | 108.6s |
| #63 | Qwen3.6 Plus medium | Qwen | 6.1 | 7.8 | $0.418 | 1/3 | 153.1s |
| #66 | GPT-5.6 Terra medium | OpenAI | 6.1 | 7.8 | $0.523 | 1/3 | 7.19s |
| #67 | Claude Sonnet 4.6 medium | Anthropic | 5.7 | 7.8 | $2.126 | 1/3 | 33.3s |
| #70 | Claude Opus 4.8 low | Anthropic | 6.6 | 7.8 | $2.089 | 1/3 | 7.58s |
| #73 | Seed-2.0-Code low | Bytedance Seed | 7.0 | 7.7 | $0.767 | 1/3 | 109.7s |
| #74 | Claude Opus 4.6 medium | Anthropic | 5.7 | 7.7 | $2.961 | 1/3 | 30.1s |
| #75 | DeepSeek V4 Pro high | DeepSeek | 6.3 | 7.7 | $0.761 | 1/3 | 243.0s |
| #78 | GPT-5.6 Luna high | OpenAI | 5.5 | 7.6 | $0.179 | 1/3 | 15.6s |
| #79 | Solar Pro 4 xhigh | Upstage | 5.8 | 7.6 | $0.050 | 1/3 | 292.2s |
| #81 | Claude Opus 5 none | Anthropic | 5.9 | 7.5 | $1.550 | 1/3 | 5.54s |
| #82 | GPT-5.4 Nano medium | OpenAI | 6.1 | 7.5 | $0.142 | 1/3 | 19.1s |