Navigate
AI BENCHY
AD
Track all your projects in one dashboard. Get 📊stats, 🔥heatmaps and 👀recordings in one self-hosted dashboard.
uxwizz.com

DeepSeek: DeepSeek V4 Flash vs MoonshotAI: Kimi K2.7 Code

DeepSeek V4 Flash (high) leads on average score with 7.7 vs 7.5. DeepSeek V4 Flash (high) has the lower benchmark cost at $0.042 vs $0.690. DeepSeek V4 Flash (high) is faster at 49.75s vs 84.25s, with pass rates of 72.7% vs 65.2%.

Recommended modelDeepSeek V4 Flash (high)It has the best score here (7.7), while costing about 16.4x less than Kimi K2.7 Code (medium).

Last updated at: 2026-07-17

Metric DeepSeek V4 Flash DeepSeek V4 Flash high Release: 2026-04-24 Kimi K2.7 Code Kimi K2.7 Code medium Release: 2026-06-12
Score 7.7 7.5
Rank #43 #50
Reliability 10.0 10.0
Consistency 8.2 8.3
Tests Correct
Attempt pass rate 72.7% 65.2%
Flaky tests 5 4
Total Runs 66 66
Cost per result 0.402 6.457
Total Cost $0.042 $0.690
Input Price $0.098 / 1M $0.750 / 1M
Output Price $0.196 / 1M $3.500 / 1M
Total Input Tokens 108,392 72,073
Output Tokens 14,478 83,714
Reasoning Tokens 153,687 178,793
Response Time (avg) 49.75s 84.25s
Response Time (max) 218.13s 365.80s
Response Time (total) 1094.41s 1769.22s

Hamster playing table tennis

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

#43 DeepSeek V4 Flash

high
Cost
$0.003
Time
93.1s
Tokens
7,926 tok

#50 MoonshotAI: Kimi K2.7 Code

medium
Cost
$0.025
Time
138.0s
Tokens
6,093 tok

Top Models by Score

Score vs Total Cost

Response Time (avg)

Score vs Response Time (avg)

Total Output Tokens

Score vs Total Output Tokens

Category Breakdown

Coding Score Consistency Attempt pass rate Flaky tests Tests Correct Response Time (avg) Input Tokens Output Tokens Reasoning Tokens
DeepSeek V4 Flash 7.8 10.0 66.7% 0 50.60s 7,279 395 34,862
Kimi K2.7 Code 7.8 9.3 66.7% 0 146.73s 4,650 1,864 25,635

Quick Compare

Switch Comparison Pair