Kategori AI BENCHY
Peringkat Kepatuhan instruksi
Lihat model AI mana yang paling baik di Kepatuhan instruksi, mana yang tetap andal, dan di mana kesenjangan terbesar muncul. Urutkan berdasarkan: Tes benar ↑.
| Peringkat | Model | Perusahaan | Skor Kepatuhan instruksi | Skor | Tes benar | Waktu respons (rata-rata) |
|---|---|---|---|---|---|---|
| #27 | Gemma 4 31B medium | 10.0 | 7.8 | 2/2 | 12.8s | |
| #28 | Gemini 2.5 Flash medium | 9.8 | 7.8 | 2/2 | 2.62s | |
| #29 | Qwen3.5-122B-A10B medium | Qwen | 10.0 | 7.8 | 2/2 | 9.88s |
| #30 | Qwen3.5-27B medium | Qwen | 10.0 | 7.8 | 2/2 | 19.7s |
| #31 | DeepSeek V4 Flash high | DeepSeek | 10.0 | 7.7 | 2/2 | 15.4s |
| #33 | Hy3 preview medium | Tencent | 10.0 | 7.7 | 2/2 | 6.16s |
| #34 | Qwen3.7 Max none | Qwen | 10.0 | 7.7 | 2/2 | 943ms |
| #35 | Gemini 3 PRO Preview medium | 9.8 | 7.6 | 2/2 | 3.26s | |
| #36 | Qwen3.5 Plus 2026-04-20 medium | Qwen | 10.0 | 7.6 | 2/2 | 20.2s |
| #37 | Gemma 4 26B A4B medium | 10.0 | 7.6 | 2/2 | 17.5s | |
| #38 | Grok 4.3 medium | X AI | 9.8 | 7.6 | 2/2 | 18.6s |
| #39 | Qwen3.6 Flash medium | Qwen | 10.0 | 7.5 | 2/2 | 6.05s |
| #40 | Gemini 3.1 Flash Lite Preview medium | 10.0 | 7.5 | 2/2 | 1.91s | |
| #41 | Nemotron 3 Ultra 550b A55b medium | NVIDIA | 9.8 | 7.5 | 2/2 | 6.35s |
| #42 | GPT-5.2 medium | OpenAI | 9.9 | 7.5 | 2/2 | 3.12s |