#1 Gemini 3.5 Flash
high- Biaya
- $0.208
- Waktu
- 118.2s
- Token
- 23,158 tok
AI BENCHY Compare
Ringkasan
Gemini 3.5 Flash (high) unggul dalam skor rata-rata dengan 9.8 vs 8.0. KAT-Coder-Pro V2.5 (low) memiliki biaya benchmark lebih rendah di $0.220 vs $1.115. Gemini 3.5 Flash (high) lebih cepat di 8.84s vs 10.44s, dengan tingkat keberhasilan 96.8% vs 71.4%.
Benchmark dihasilkan dari suite pengujian AI BENCHY pada: 2026-07-14
| Metrik | Gemini 3.5 Flash Gemini 3.5 Flash high | KAT-Coder-Pro V2.5 KAT-Coder-Pro V2.5 low |
|---|---|---|
| Skor | 9.8 | 8.0 |
| Peringkat | #1 | #35 |
| Keandalan | 10.0 | 9.9 |
| Konsistensi | 9.6 | 7.3 |
| Tes benar | ||
| Tingkat lulus per percobaan | 96.8% | 71.4% |
| Tes tidak stabil | 1 | 7 |
| Total Run | 63 | 63 |
| Biaya per hasil | 5.575 | 1.991 |
| Total Biaya | $1.115 | $0.220 |
| Harga input | $1.500 / 1M | $0.740 / 1M |
| Harga output | $9.000 / 1M | $2.960 / 1M |
| Total token input | 37,594 | 42,750 |
| Token output | 1,975 | 2,167 |
| Token penalaran | 115,638 | 61,133 |
| Waktu respons (rata-rata) | 8.84s | 10.44s |
| Waktu respons (maks) | 34.82s | 51.34s |
| Waktu respons (total) | 185.57s | 219.17s |
Showcase generasi
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
| Pemrograman | Skor | Konsistensi | Tingkat lulus per percobaan | Tes tidak stabil | Tes benar | Waktu respons (rata-rata) | Token input | Token output | Token penalaran |
|---|---|---|---|---|---|---|---|---|---|
| Gemini 3.5 Flash | 10.0 | 10.0 | 100.0% | 0 | 22.96s | 8,118 | 456 | 47,129 | |
| KAT-Coder-Pro V2.5 | 7.8 | 10.0 | 66.7% | 0 | 24.87s | 7,893 | 402 | 24,945 |
Ganti Pasangan Perbandingan