Model comparison

GPT-5.2 vs Qwen2.5 72B Instruct

GPT-5.2 is the stronger model overall, scoring 54.1 to 31.9 on the Noometry Index. Qwen2.5 72B Instruct costs 2.0× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.

Last verified October 10, 2026. 25 shared benchmarks.

GPT-5.2 OpenAI

54.1

Rank #34 Confirmed

Qwen2.5 72B Instruct

Alibaba (Qwen)

31.9

Rank #267 Confirmed

Summary

They share 25 benchmarks with published results for both. GPT-5.2 scores higher in 9 categories and Qwen2.5 72B Instruct in 0 categories; 9 gaps are clear of the uncertainty.
The widest gap is in math, where GPT-5.2 leads 60.0 to 19.3.
The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 96.1% for GPT-5.2 and 8.1% for Qwen2.5 72B Instruct.
Qwen2.5 72B Instruct is cheaper at $1.40 / $5.60 per million input/output tokens, against $1.75 / $14 for GPT-5.2.
GPT-5.2 accepts more context: 400K tokens versus 131K.
Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.

Side by side

GPT-5.2 and Qwen2.5 72B Instruct specifications
	GPT-5.2	Qwen2.5 72B Instruct
Provider	OpenAI	Alibaba (Qwen)
Noometry Index	54.1	31.9
Released	2025-12-11	2024-09
Weights	Proprietary	Open
Context window	400K	131K
Max output	128K	8K
Input $ / M tokens	$1.75	$1.40
Output $ / M tokens	$14	$5.60
Results tracked	67	43

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding GPT-5.2 leads

GPT-5.2: 51.6 (#37), Qwen2.5 72B Instruct: 33.2 (#260)

Coding benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
WeirdML	72.2%	16%
LMArena Coding	1447	1292
SWE-bench Verified	73.8%	—
SWE-bench Verified (bash only)	72.8%	—
LMArena WebDev	1416	—
SWE-bench Multilingual	66.7%	—
GSO	27.4%	—
BigCodeBench Instruct	—	45.8%
BigCodeBench Complete	—	55.9%
ALE-Bench	1,294	—
AlgoTune	2.05	—

Agentic & Tool Use GPT-5.2 leads

GPT-5.2: 40.2 (#24), Qwen2.5 72B Instruct: 22.1 (#133)

Agentic & Tool Use benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
METR Time Horizons	75.3%	35.8%
Terminal-Bench	64.9%	—
Berkeley Function Calling Leaderboard	55.9%	—
GDPval	49.7%	—
Remote Labor Index	2.5%	—
TheAgentCompany	—	5.7%
τ²-bench Airline	83%	—
τ²-bench Banking	32.2%	—
τ²-bench Retail	81.6%	—
τ²-bench Telecom	89.7%	—
DeepResearch Bench	41.1%	—
BALROG	—	16.2%
LMArena Search	1207	—
Vending-Bench 2	3,591	—

Reasoning GPT-5.2 leads

GPT-5.2: 50.2 (#35), Qwen2.5 72B Instruct: 22.3 (#199)

Reasoning benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
LMArena Hard Prompts	1445	1271
DTBench	90.9%	62.9%
LMCA	43.9%	13.4%
Epoch Capabilities Index	153.45	129
ForecastBench	60.1	57.5
ARC-AGI-2	52.9%	—
SimpleBench	45.8%	—
Kagi LLM Benchmark	73.3%	—
NYT Connections (extended)	83.6%	—
ARC-AGI-1	86.2%	—
Chess Puzzles	49%	—
EnigmaEval	10.4%	—
EBR-Bench	23%	—
Mystery Game Puzzles	23%	—
BIG-Bench Hard	—	79.8%
HellaSwag	—	84.8%
PIQA	—	82.6%
WinoGrande	—	82.3%

Math GPT-5.2 leads

GPT-5.2: 60.0 (#38), Qwen2.5 72B Instruct: 19.3 (#287)

Math benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
OTIS Mock AIME 2024-2025	96.1%	8.1%
LMArena Math	1440	1283
FrontierMath (Tiers 1-3)	67.4%	—
FrontierMath Tier 4	31.7%	—
MathArena Final-Answer Competitions	72%	—
ProofBench	15%	—
Omni-MATH	—	33%
MATH Level 5	—	63.2%
FrontierMath (Feb 2025 set)	40.7%	—
FrontierMath Tier 4 (v1)	18.8%	—

Knowledge GPT-5.2 leads

GPT-5.2: 59.3 (#32), Qwen2.5 72B Instruct: 27.0 (#253)

Knowledge benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
GPQA Diamond	91.4%	49.1%
LMArena Expert	1445	1245
Humanity's Last Exam	27.8%	—
SimpleQA Verified	37.1%	—
MMLU-Pro	—	63.1%
Confabulations	—	19.1%
Vectara Hallucination Rate	8.4%	—
GPQA (HELM)	—	42.6%
ARC (AI2) Challenge	—	94.5%
MMLU	—	85.3%
TriviaQA	—	71.9%

Multimodal Not comparable

GPT-5.2: 51.3 (#7), Qwen2.5 72B Instruct: —

Multimodal benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
LMArena Vision	1268	—
VPCT	84%	—
Furniture Assembly	38.3%	—
LMArena Document	1405	—

Multilingual GPT-5.2 leads

GPT-5.2: 53.4 (#67), Qwen2.5 72B Instruct: 41.0 (#213)

Multilingual benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
LMArena Non-English	1425	1252
LMArena Chinese	1460	1272
LMArena French	1455	1280
LMArena German	1448	1234
LMArena Japanese	1420	1180
LMArena Korean	1392	1188
LMArena Russian	1440	1264
LMArena Spanish	1433	1256

Instruction Following GPT-5.2 leads

GPT-5.2: 74.7 (#89), Qwen2.5 72B Instruct: 65.5 (#221)

Instruction Following benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
LMArena Instruction Following	1417	1254
IFEval	—	80.6%

Long Context GPT-5.2 leads

GPT-5.2: 44.0 (#78), Qwen2.5 72B Instruct: 38.9 (#188)

Long Context benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
LMArena Longer Query	1428	1282
CL-bench	18.2%	—

Writing & Preference GPT-5.2 leads

GPT-5.2: 66.8 (#32), Qwen2.5 72B Instruct: 46.7 (#215)

Writing & Preference benchmarks
Benchmark	GPT-5.2	Qwen2.5 72B Instruct
LMArena Text	1439	1269
LMArena Creative Writing	1401	1221
LMArena Multi-Turn	1458	1272
EQ-Bench Creative Writing	1703	—
WildBench	—	80.2%

Frequently asked questions

Is GPT-5.2 better than Qwen2.5 72B Instruct?

Which is cheaper, GPT-5.2 or Qwen2.5 72B Instruct?

Qwen2.5 72B Instruct is cheaper. It lists at $1.40 per million input tokens and $5.60 per million output tokens; GPT-5.2 lists at $1.75 and $14.

Is GPT-5.2 or Qwen2.5 72B Instruct better for coding?

GPT-5.2 scores higher on coding benchmarks: 51.6 versus 33.2 in the Noometry coding category.

Which has the bigger context window?

GPT-5.2 does, with 400K tokens against 131K.

How many benchmarks do GPT-5.2 and Qwen2.5 72B Instruct share?

25 benchmarks have published results for both models. GPT-5.2 has 67 scored results on Noometry and Qwen2.5 72B Instruct has 43.