Navigasi
AI BENCHY
Advertise here

DeepSeek: DeepSeek V4 Flash vs Google: Gemini 2.5 Flash

Gemini 2.5 Flash (medium) unggul dalam skor rata-rata dengan 8.2 vs 7.7. DeepSeek V4 Flash (high) memiliki biaya benchmark lebih rendah di $0.041 vs $0.643. Gemini 2.5 Flash (medium) lebih cepat di 21.18s vs 49.75s, dengan tingkat keberhasilan 72.7% vs 71.2%.

Model yang direkomendasikanDeepSeek V4 Flash (high)Its score stays close to the best score here (7.7 vs 8.2), while costing about 16.0x less than Gemini 2.5 Flash (medium).

Benchmark dihasilkan dari suite pengujian AI BENCHY pada: 2026-07-25

Metrik DeepSeek V4 Flash DeepSeek V4 Flash high Rilis: 2026-04-24 Gemini 2.5 Flash Gemini 2.5 Flash medium Rilis: 2025-06-17
Skor 7.7 8.2
Peringkat #52 #31
Keandalan 10.0 10.0
Konsistensi 8.2 9.6
Tes benar
Tingkat lulus per percobaan 72.7% 71.2%
Tes tidak stabil 5 1
Total Run 66 66
Biaya per hasil 0.402 4.286
Total Biaya $0.041 $0.643
Harga input $0.094 / 1M $0.300 / 1M
Harga output $0.188 / 1M $2.500 / 1M
Total token input 108,392 132,498
Token output 14,478 12,739
Token penalaran 153,687 228,464
Waktu respons (rata-rata) 49.75s 21.18s
Waktu respons (maks) 218.13s 140.50s
Waktu respons (total) 1094.41s 465.89s

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#52 DeepSeek V4 Flash

high
Biaya
$0.003
Waktu
93.1s
Token
7,926 tok

#31 Gemini 2.5 Flash

medium
SVG tidak valid
Biaya
$0.000
Waktu
274.0s
Token
0 tok

Model teratas berdasarkan skor

Skor vs Total Biaya

Waktu respons (rata-rata)

Skor vs Waktu respons (rata-rata)

Total token output

Skor vs Total token output

Rincian Kategori

Pemrograman Skor Konsistensi Tingkat lulus per percobaan Tes tidak stabil Tes benar Waktu respons (rata-rata) Token input Token output Token penalaran
DeepSeek V4 Flash 7.8 10.0 66.7% 0 50.60s 7,279 395 34,862
Gemini 2.5 Flash 7.8 10.0 66.7% 0 41.01s 6,669 543 32,303

Perbandingan Cepat

Ganti Pasangan Perbandingan

Claude Opus 4.6mediumvsDeepSeek V4 FlashhighGemini 2.5 FlashmediumvsMuse Spark 1.1lowDeepSeek V4 FlashhighvsGLM 5mediumDeepSeek V4 FlashhighvsMiniMax M3mediumDeepSeek V4 FlashhighvsGrok Build 0.1mediumClaude Opus 4.8lowvsDeepSeek V4 FlashhighDeepSeek V4 FlashhighvsGPT-5.6 LunamediumDeepSeek V4 FlashhighvsNemotron 3 UltramediumTersedia gratisClaude Opus 5nonevsDeepSeek V4 FlashhighDeepSeek V4 FlashhighvsKimi K2.7 CodemediumDeepSeek V4 FlashhighvsGPT-5.4 NanomediumGemini 2.5 FlashmediumvsMuse Spark 1.1high