AI BENCHY
Advertise here
#57

GPT-5.2

OpenAI Release: 2025-12-11 Tested on: 2026-05-22 00:19 openai/gpt-5.2::medium
~1.2T total (~80B active)MoEClosedEstimated

Summary

GPT-5.2 scores 7.4 on AI BENCHY and ranks #57. It has 10.0 reliability, a 70.0% pass rate, $0.490 total cost, and 16.44s average response time.

What makes GPT-5.2 unique: It stands out most in Coding, where it ranks #1, while Anti-AI Tricks is its weakest area at #18.

Model facts

Researched on 2026-08-12

Estimated
Parameters
~1.2T total (~80B active)
Architecture
MoE
Availability
Closed
License
-

Best estimate from public evidence; the vendor did not disclose every value. Vendor does not disclose the architecture or count.

Parameter source Availability source

Score

7.4

Consistency

8.3

Reliability

10.0

Total Cost (Current Price)

$0.490

Total Output Tokens

31,057

Total Input Tokens

0

Input Price

$1.750 / 1M

Output Price

$14.000 / 1M

Tests Correct

Wrong Tests: 8

Attempt pass rate: 70.0%

Flaky tests

4

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

16.44s

Response Time (max): 77.80s

Response Time (total): 213.77s

Did not follow instructions: 3 Wrong answer: 3 No answer: 1 Timed out: 1

Hamster playing table tennis

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

#57 GPT-5.2

medium
Cost
$0.047
Time
49.2s
Tokens
3,396 tok

Run history

Tested on Score Reliability Tests Correct Total Cost Compare
2026-08-14 02:17 Re-test 8.4 10.0 $1.182 Compare
2026-07-16 22:37 New test added 8.4 10.0 $0.951 Compare
2026-06-04 13:45 New test added 7.5 10.0 $0.548 Compare
2026-05-22 00:19 Re-test 7.4 10.0 $0.490 Current run
2026-04-11 01:44 First recorded run 7.5 N/A $0.352 Compare

Run comparison

RunBenchmark coverageScoreConsistencyReliabilityTests CorrectFlaky testsTotal Output TokensTotal Input TokensTotal CostResponse Time (avg)
2026-05-22 00:19 · Re-test60/60 attempts7.48.310.012/20431,0570$0.49016.44s
2026-07-16 22:37 · New test added66/66 attempts8.48.510.014/22454,782105,004$0.95122.62s
Difference—-1.0-0.20.0-20-23725-105004-$0.462-6174ms

Benchmark coverage differs: 60/60 attempts (Target: 3 repeats per test) versus 66/66 attempts (Target: 3 repeats per test). Totals and repeat-sensitive metrics are not directly comparable.

These two runs used different benchmark suites, so the deltas reflect both model changes and suite changes.

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

GPT-5.2mediumvsGrok 4.5lowGPT-5.2mediumvsQwen3.8 27BhighGPT-5.2mediumvsGrok 4.6mediumGPT-5.2mediumvsSeed 2.1 TurbomediumGPT-5.2mediumvsMuse Spark 1.2lowGPT-5.2mediumvsInkling SmallhighGPT-5.2mediumvsGrok 4.5mediumGPT-5.2mediumvsMuse Spark 1.1lowGPT-5.2mediumvsGPT-5.4mediumGPT-5.2mediumvsGemini 2.5 Flashmedium

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 6.5 8.0
Coding 10.0 10.0
Combined 10.0 10.0
Data parsing and extraction 10.0 10.0
Domain specific 5.9 7.2
General Intelligence 3.7 9.7
Instructions following 9.9 10.0
Puzzle Solving 7.6 7.3
Tool Calling 4.7 1.6
Trivia 3.0 10.0

Compared models