GPT-5.2

Released	2025-12-11
Price per 1M	$1.75 in · $14.00 out
Overall score	56.2 · rank 65
Context window	400,000 tokens
Provider	OpenAI
Weights	Proprietary

Scores

Dimension	Score	Rank	Benchmarks
Overall: Balanced	56.2	65	41
Overall: Coding	55.9	63	26
Overall: Reasoning	57.2	61	17
Overall: Agentic work	55.6	75	24
Overall: Frontend	54.4	62	10
Coding	56.8	59	5
Agents	54.3	78	10
Reasoning	58.7	60	9
Math	57.4	50	5
Knowledge	53.5	73	3
Frontend	52.4	63	2
General	57.0	46	4
Multimodal	59.1	26	3
Value vs price	+3.8	88

Configurations

Configuration	Balanced	Benchmarks
Unspecified setting	52.1 ±2.6	16
High	55.6 ±1.1	22
Low	49.9 ±3.7	11
Medium	55.5 ±2.5	13
Non-reasoning	45.4 ±4.3	10
Extra High	56.9 ±2.1	18

Evidence

Coding 56.8 ±1.6

Benchmark	Result	Source	Date
WeirdML v3	5.0%	Epoch AI	2026-11-11
MirrorCode	32.2%	Epoch AI	2026-10-21
GSO	27.5%	Epoch AI	2026-11-11
LiveCodeBench (Vals run)	85.4%	Vals AI	2026-09-01
SWE-bench Verified (Epoch run)	73.8%	Epoch AI	2026-02-12
SWE-bench Verified (Vals run)	75.8%	Vals AI	2026-09-01

Agents 54.3 ±2.6

Benchmark	Result	Source	Date
Terminal-Bench 2.0	64.9%	Epoch AI	2025-12-11
Remote Labor Index	2.5%	Epoch AI	2026-11-11
METR 50% Time Horizon	$352	METR	2026-11-11
tau2-bench Banking Knowledge	32.2%	tau2-bench	2026-05-05
BFCL v4	55.9%	BFCL	2026-04-12
METR task suite	75.3%	Epoch AI	2026-11-11
tau2-bench Airline	83.0%	tau2-bench	2026-05-05
DeepResearch Bench	41.1%	Epoch AI	2026-11-11
Vending-Bench 2	$3,591	Epoch AI	2026-11-11
tau2-bench Telecom	89.7%	tau2-bench	2026-05-05
tau2-bench Retail	81.6%	tau2-bench	2026-05-05

Reasoning 58.7 ±1.7

Benchmark	Result	Source	Date
ARC-AGI-2	52.9%	Epoch AI	2026-11-11
Extended NYT Connections	83.6%	Lech Mazur	2026-11-10
ARC-AGI-2 (ARC Prize verified)	52.9%	ARC Prize	2026-11-11
ARC-AGI-1 (ARC Prize verified)	86.2%	ARC Prize	2026-11-11
ARC-AGI-1	86.2%	Epoch AI	2026-11-11
EnigmaEval	10.4%	Epoch AI	2026-11-11
GPQA Diamond	91.4%	Epoch AI	2025-12-13
GPQA Diamond (Vals run)	91.7%	Vals AI	2026-09-01
Kagi Reasoning Benchmark	73.3%	Kagi	2026-11-11
Chess Puzzles	48.0%	Epoch AI	2026-10-14
SimpleBench	45.8%	SimpleBench	2025-12-12
SimpleBench	45.8%	Epoch AI	2026-11-11
Mystery Game Puzzles	23.0%	Epoch AI	2026-08-06

Math 57.4 ±1.6

Benchmark	Result	Source	Date
FrontierMath Tiers 1-3	67.4%	Epoch AI	2026-06-11
FrontierMath Tier 4	31.7%	Epoch AI	2026-06-11
ProofBench	15.0%	Epoch AI	2026-11-11
OTIS Mock AIME 2024-2025	96.1%	Epoch AI	2025-12-11
MathArena Final-Answer Competitions	72.0%	MathArena	2026-11-11
FrontierMath Tiers 1-3 (2025 set)	40.7%	Epoch AI	2025-12-13
FrontierMath Tier 4 (2025 set)	18.8%	Epoch AI	2025-12-14

Knowledge 53.5 ±2.6

Benchmark	Result	Source	Date
SimpleQA Verified	37.1%	Epoch AI	2026-08-27
Humanity's Last Exam	27.8%	Epoch AI	2026-11-11
MMLU-Pro (Vals run)	86.2%	Vals AI	2026-09-01

Frontend 52.4 ±0.7

Benchmark	Result	Source	Date
Vibe Code Bench v1.1	53.5%	Vals AI	2026-11-09
LMArena WebDev	1,420	LMArena	2026-11-09

General 57.0 ±2.2

Benchmark	Result	Source	Date
EQ-Bench Creative Writing v3	1,703	EQ-Bench	2026-11-11
LMArena Text	1,437	LMArena	2026-11-09
Creative Story-Writing (Lech Mazur)	54.0%	Lech Mazur	2026-10-23
EQ-Bench Longform Creative Writing	71.6%	EQ-Bench	2026-11-11

Multimodal 59.1 ±4.8

Benchmark	Result	Source	Date
LMArena Vision	1,242	LMArena	2026-11-09
MMMU-Pro (Vals run)	86.7%	Vals AI	2026-09-01
VPCT	84.0%	Epoch AI	2026-11-11

Results not used in the score

Benchmark	Result	Source	Why not used
ARC-AGI-1	12.3%	Epoch AI	A stronger configuration result was used
ARC-AGI-1	12.3%	Epoch AI	A stronger configuration result was used
ARC-AGI-1	78.7%	Epoch AI	A stronger configuration result was used
ARC-AGI-1	72.7%	Epoch AI	A stronger configuration result was used
ARC-AGI-1	55.7%	Epoch AI	A stronger configuration result was used
ARC-AGI-1 (ARC Prize verified)	12.3%	ARC Prize	A stronger configuration result was used
ARC-AGI-1 (ARC Prize verified)	78.7%	ARC Prize	A stronger configuration result was used
ARC-AGI-1 (ARC Prize verified)	72.7%	ARC Prize	A stronger configuration result was used
ARC-AGI-1 (ARC Prize verified)	55.7%	ARC Prize	A stronger configuration result was used
ARC-AGI-2	0.8%	Epoch AI	A stronger configuration result was used
ARC-AGI-2	43.3%	Epoch AI	A stronger configuration result was used
ARC-AGI-2	26.7%	Epoch AI	A stronger configuration result was used
ARC-AGI-2	9.7%	Epoch AI	A stronger configuration result was used
ARC-AGI-2 (ARC Prize verified)	0.8%	ARC Prize	A stronger configuration result was used
ARC-AGI-2 (ARC Prize verified)	43.3%	ARC Prize	A stronger configuration result was used
ARC-AGI-2 (ARC Prize verified)	26.7%	ARC Prize	A stronger configuration result was used
ARC-AGI-2 (ARC Prize verified)	9.7%	ARC Prize	A stronger configuration result was used
BFCL v4	45.3%	BFCL	A stronger configuration result was used
Chess Puzzles	2.0%	Epoch AI	A stronger configuration result was used
Chess Puzzles	45.0%	Epoch AI	A stronger configuration result was used
Chess Puzzles	36.0%	Epoch AI	A stronger configuration result was used
Chess Puzzles	26.0%	Epoch AI	A stronger configuration result was used
Extended NYT Connections	14.5%	Lech Mazur	A stronger configuration result was used
Extended NYT Connections	68.1%	Lech Mazur	A stronger configuration result was used
Extended NYT Connections	60.6%	Lech Mazur	A stronger configuration result was used
Extended NYT Connections	54.2%	Lech Mazur	A stronger configuration result was used
FrontierMath Tier 4 (2025 set)	18.8%	Epoch AI	A stronger configuration result was used
FrontierMath Tier 4 (2025 set)	16.7%	Epoch AI	A stronger configuration result was used
FrontierMath Tier 4 (2025 set)	6.3%	Epoch AI	A stronger configuration result was used
FrontierMath Tiers 1-3 (2025 set)	40.3%	Epoch AI	A stronger configuration result was used
FrontierMath Tiers 1-3 (2025 set)	36.9%	Epoch AI	A stronger configuration result was used
FrontierMath Tiers 1-3 (2025 set)	26.6%	Epoch AI	A stronger configuration result was used
GPQA Diamond	73.2%	Epoch AI	A stronger configuration result was used
GPQA Diamond	88.2%	Epoch AI	A stronger configuration result was used
GPQA Diamond	87.9%	Epoch AI	A stronger configuration result was used
GPQA Diamond	82.7%	Epoch AI	A stronger configuration result was used
LMArena Text	1,435	LMArena	A stronger configuration result was used
LMArena Vision	1,228	LMArena	A stronger configuration result was used
Mystery Game Puzzles	14.0%	Epoch AI	A stronger configuration result was used
Mystery Game Puzzles	22.0%	Epoch AI	A stronger configuration result was used
Mystery Game Puzzles	10.0%	Epoch AI	A stronger configuration result was used
OTIS Mock AIME 2024-2025	62.2%	Epoch AI	A stronger configuration result was used
OTIS Mock AIME 2024-2025	96.1%	Epoch AI	A stronger configuration result was used
OTIS Mock AIME 2024-2025	93.9%	Epoch AI	A stronger configuration result was used
OTIS Mock AIME 2024-2025	78.9%	Epoch AI	A stronger configuration result was used
Remote Labor Index	2.1%	Epoch AI	A stronger configuration result was used
SimpleBench	45.8%	Epoch AI	A stronger configuration result was used
SimpleQA Verified	34.3%	Epoch AI	A stronger configuration result was used
SimpleQA Verified	32.7%	Epoch AI	A stronger configuration result was used
SimpleQA Verified	32.8%	Epoch AI	A stronger configuration result was used
Terminal-Bench 2.0	62.9%	Epoch AI	A stronger configuration result was used
Terminal-Bench 2.0	60.7%	Epoch AI	A stronger configuration result was used
Terminal-Bench 2.0	54.0%	Epoch AI	A stronger configuration result was used
Terminal-Bench 2.0	64.9%	Epoch AI	A stronger configuration result was used
Terminal-Bench 2.0	62.9%	Epoch AI	A stronger configuration result was used
VPCT	67.0%	Epoch AI	A stronger configuration result was used
tau2-bench Airline	52.5%	tau2-bench	A stronger configuration result was used
tau2-bench Banking Knowledge	12.6%	tau2-bench	A stronger configuration result was used
tau2-bench Retail	75.0%	tau2-bench	A stronger configuration result was used
tau2-bench Telecom	57.2%	tau2-bench	A stronger configuration result was used

Names matched to this model

gpt-5-2, gpt-5-2-2025-12-11, robin-alpha-next-2025-11-24

Compare GPT-5.2 with other models