Navigate
AI BENCHY
Advertise here

Gemma 4 31B (medium) vs GPT-5.6 Luna (low)

The average score is effectively tied at 6.2 vs 6.2. GPT-5.6 Luna (low) has the lower benchmark cost at $0.051 vs $0.100. GPT-5.6 Luna (low) is faster at 5.20s vs 78.59s, with pass rates of 65.2% vs 54.5%.

Last updated at: 2026-08-20

Rank
#166
Total Output Tokens
267,421
Response Time (avg)
78.59s
Total Cost
$0.100
Rank
#167
Total Output Tokens
26,286
Response Time (avg)
5.20s
Total Cost
$0.051
Recommended model GPT-5.6 Luna (low)

It has the best score here (6.2), while costing about 2.0x less than Gemma 4 31B (medium).

Detailed comparison

Metric Gemma 4 31B Gemma 4 31B medium Release: 2026-04-02 Free Available GPT-5.6 Luna GPT-5.6 Luna low Release: 2026-07-09
Score 6.2 6.2
Rank #166 #167
Reliability 10.0 10.0
Consistency 8.7 8.2
Attempts 66/66 66/66
Tests Correct
Attempt pass rate 65.2% 54.5%
Flaky tests 3 5
Total Runs 66 66
Cost per result 1.152 2.437
Total Cost $0.100 $0.051
Input Price $0.090 / 1M $0.200 / 1M
Output Price $0.340 / 1M $1.200 / 1M
Total Input Tokens 94,969 96,355
Output Tokens 34,399 8,213
Reasoning Tokens 233,022 18,073
Response Time (avg) 78.59s 5.20s
Response Time (max) 437.40s 19.44s
Response Time (total) 1571.84s 114.46s
Parameters 30.7B ~400B total (~17B active)
Availability Open source Closed

Model generation showcase

Hamster playing table tennis

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

#166 Gemma 4 31B

medium
Cost
$0.002
Time
45.7s
Tokens
2,696 tok

#167 GPT-5.6 Luna

low
Cost
$0.011
Time
10.2s
Tokens
1,897 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
Gemma 4 31B 4.3 5.8 22.2% 1 219.76s 5,568 11,098 33,212
GPT-5.6 Luna 5.5 10.0 33.3% 0 4.61s 7,302 557 3,535

Quick Compare

Switch Comparison Pair