modelbenchmark.io

GPT-5.2

OpenAI · gpt-5-2

Compare

Reliable GPT generation for broad coding, writing, and tool-assisted product work

Specification

openai listing

Context
400K
Max output
128K
Released
2025-12-11
Knowledge cutoff
2025-08-31
Retires
2027-06-08
Open weights
no
Input
text, image
Output
text

Price

US dollars per million tokens · openai list price

Input
$1.75
Output
$14.00
Cache read
$0.175
Cache write
$1.75

Batch and priority prices

US dollars per million tokens · per listing

PriceListing$/M
Batch inputopenai$0.875
Batch outputopenai$7.00
Priority inputazure$3.50
Priority outputazure$28.00

Batch input is 50% below the interactive input price at the same listing.

Quality

12 benchmarks · 3 sources

Composite
79th
percentile of 334 scored models
Rank
70 / 334
±0.06 sd
Evidence
12 × 3
3 or more benchmarks, from 2 or more sources
Effort range
14.3
points between effort settings
coding73rd1/3 bench
reasoning84th4/4 bench
math82nd5/7 bench
knowledge34th1/1 bench
general45th1/1 bench

By reasoning effort

every setting placed on the same scale as the leaderboard

SettingCompositeBenchmarks behind it
xhigh86th9
high84th9
medium84th7
low63rd7

Every score

one row per source and configuration — nothing averaged away

BenchmarkScoreConfigurationSourceRun
Chess Puzzles49.0 ±5.0xhighEpoch AI2025-12-15
40.0 ±4.9highEpoch AI2025-12-11
40.0 ±4.9mediumEpoch AI2025-12-11
23.0 ±4.2lowEpoch AI2025-12-11
Settings differ by 26.0 points on this benchmark.
OTIS Mock AIME 2024-202596.1 ±2.6highEpoch AI2025-12-11
96.1 ±2.7xhighEpoch AI2025-12-13
93.9 ±3.1mediumEpoch AI2025-12-11
78.9 ±5.0lowEpoch AI2025-12-11
Settings differ by 17.2 points on this benchmark.
GPQA diamond91.4 ±1.8xhighEpoch AI2025-12-13
88.2 ±1.9highEpoch AI2025-12-11
87.9 ±1.9mediumEpoch AI2025-12-11
82.7 ±2.3lowEpoch AI2025-12-11
Settings differ by 8.7 points on this benchmark.
FrontierMath-v150.2 ±2.9high · 2 runsEpoch AI2025-12-11
48.5 ±16.3medium · 2 runsEpoch AI2025-12-11
40.7 ±2.9xhighEpoch AI2025-12-13
33.3 ±2.6low · 2 runsEpoch AI2025-12-11
Settings differ by 16.9 points on this benchmark.
FrontierMath-Tiers-1-3-v267.4 ±2.8xhighEpoch AI2026-06-11
SWE-Bench verified73.8 ±2.0highEpoch AI2026-02-12
72.3high · mini-SWE-agent · 2 runsSWE-bench
69.0default · mini-SWE-agentSWE-bench
Sources and settings disagree by 4.8 points. Both figures stand.
Mystery Game Puzzles23.0 ±4.2highEpoch AI2026-08-06
22.0 ±4.2mediumEpoch AI2026-08-28
10.0 ±3.0lowEpoch AI2026-08-27
Settings differ by 13.0 points on this benchmark.
LiveBench75.2high · 23 runsLiveBench2026-06-25
FrontierMath-Tier-4-2025-07-0118.8 ±5.6xhighEpoch AI2025-12-14
9.4 ±0.0high · 2 runsEpoch AI2025-12-11
8.3 ±5.4medium · 2 runsEpoch AI2025-12-11
3.1 ±0.0low · 2 runsEpoch AI2025-12-11
Settings differ by 15.7 points on this benchmark.
FrontierMath-Tier-4-v231.7 ±7.3xhighEpoch AI2026-06-11
EBR-bench23.0xhighEpoch AI2026-06-26
SimpleQA Verified37.1 ±1.5xhighEpoch AI2026-08-27
34.3 ±1.5highEpoch AI2026-08-27
32.8 ±1.5lowEpoch AI2026-08-27
32.7 ±1.5mediumEpoch AI2026-08-27
Settings differ by 4.4 points on this benchmark.

The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.

Available from 34 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
openaivendor$1.75$14.00$0.175
400K
272K
128K
qihang-ai$0.25$2.00400K128K
unorouter$1.05$8.40400K128K
sap-ai-core$1.25$9.44$0.12400K128K
jiekou$1.575$12.60400K128K
poe$1.60$13.00$0.16400K128K
302ai$1.75$14.00400K128K
abacus$1.75$14.00$0.175400K128K
aihubmix$1.75$14.00$0.175400K128K
azure$1.75$14.00
$0.125
$0.175
400K
272K
128K2027-06-08
azure-cognitive-services$1.75$14.00$0.125400K128K
databricks$1.75$14.00$0.175$1.75
400K
272K
128K
edenai$1.75$14.00$0.175400K128K
gmi · gmi$1.75$14.00410K32K
impossibl$1.75$14.00$0.175400K128K
kilo$1.75$14.00$0.175400K128K
llmgateway$1.75$14.00$0.175400K128K
llmgateway-providers$1.75$14.00$0.175400K128K
merge-gateway$1.75$14.00$0.175400K128K
nano-gpt$1.75$14.00$0.175400K128K
nearai$1.75$14.00$0.175400K128K
neon$1.75$14.00$0.175400K128K
ofox$1.75$14.00$0.18400K128K
opencode$1.75$14.00$0.175400K128K
openrouter$1.75$14.00$0.175
400K
272K
128K
orcarouter$1.75$14.00$0.175400K128K
perplexity$1.75$14.00$0.175
perplexity-agent$1.75$14.00$0.175400K128K
vercel$1.75$14.00$0.175400K128K
zenmux$1.75$14.00$0.17400K64K
anyapi400K128K
chatgpt · chatgpt128K64K
github_copilot · github-copilot128K64K
qiniu-ai400K128K

Price history

append-only observations · a listing writes a row only when its price moves

DateHostField$/M
2026-09-02nearaiCache read$0.18
2026-09-10nearaiCache read$0.175
2026-09-02nearaiInput$1.80
2026-09-10nearaiInput$1.75
2026-09-02nearaiOutput$15.50
2026-09-10nearaiOutput$14.00

Other listings of this model

same model, different key

More from OpenAI

most-hosted first