GPT-5 mini

GPT-5 mini ranks first on 3 of the 160 benchmarks it has scores on.

OpenAI Released 7 Aug 2025

170

Benchmarks with a score

Ranked first, out of five or more models

In the top three

OpenAI Aug 2025

GPT-5 mini

Closed weights

145.4 #91 of 271

Epoch Capabilities Index

Humanity's Last Exam 19.4% #21 /47
SWE-bench Verified 64.7% #26 /32
Terminal-Bench 2.0 34.8% #29 /43

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

Benchmark	Score	Rank	Source
MMBench-CN v1.1	86.1%	#10 of 178	Official leaderboard
COLLIE	98.5%	#2 of 12	Model card
SimpleQA Hallucination Rate	26%	#2 of 12	Model card
Slop Score	26.3	#3 of 17	Official leaderboard
Lech Mazur Creative Writing	8.31	#8 of 44	Epoch AI
OpenAI MRCR	84.3%	#2 of 10	Model card
LongFact	0.70%	#1 of 5	Model card
Phare	73.2%	#13 of 62	Official leaderboard
SuperGPQA	58.8%	#13 of 58	Official leaderboard
LLM Confabulation Leaderboard	13.3%	#14 of 58	Official leaderboard
Wer wird Millionär? LLM Benchmark	97889.0	#19 of 76	Official leaderboard
Internal API instruction following (OpenAI)	65.8%	#1 of 4	Model card
LongFact-Concepts	0.90%	#1 of 4	Model card
FACTS Multimodal	41.5%	#12 of 39	Official leaderboard
Fiction.LiveBench	69.4%	#19 of 57	Epoch AI
LMArena (Elo)	1,390	#141 of 333	Official leaderboard
Humanity's Last Exam	19.4%	#21 of 47	Official leaderboard
MultiChallenge	59%	#17 of 35	Official leaderboard
FActScore	3.50%	#2 of 4	Model card
LongFact-Objects	1.80%	#2 of 4	Model card
FACTS Search	67.9%	#16 of 31	Official leaderboard
GraphWalks	73.4%	#12 of 21	Model card
FACTS Benchmark Suite	41.5%	#23 of 40	Official leaderboard
SimpleQA	22%	#36 of 59	Model card
Longform Creative Writing (EQ-Bench)	56.2	#75 of 119	Official leaderboard
MultiNRC	23.9%	#25 of 38	Official leaderboard
BrowseComp Long Context 128k	89.4%	#2 of 3	Model card
Vectara Hallucination Leaderboard (HHEM)	3.20%	#100 of 148	Official leaderboard
Creative Writing v3	60.0	#84 of 120	Official leaderboard
FACTS Parametric	16.1%	#35 of 45	Official leaderboard
SkateBench v2	38%	#17 of 21	Official leaderboard
Vectara Hallucination Leaderboard (2025 dataset)	12.9%	#86 of 104	Official leaderboard
SimpleQA Verified	21.6%	#70 of 84	Epoch AI
HealthBench Hard Hallucinations	3.80%	#2 of 2	Model card

Math and science

Benchmark	Score	Rank	Source
MATH Level 5	97.8%	#2 of 94	Epoch AI
SMT 2025	89.2%	#11 of 40	Official leaderboard
AIME 2025	91.1%	#44 of 143	Model card
FrontierMath	27.2%	#22 of 70	Epoch AI
OTIS Mock AIME 2024-2025	86.7%	#70 of 202	Epoch AI
FrontierMath Tier 4	6.25%	#24 of 55	Epoch AI
CMIMC 2025	84.4%	#15 of 33	Official leaderboard
GPQA Diamond	75%	#102 of 219	Epoch AI
DTBench	67.5%	#86 of 178	Official leaderboard
HMMT 2025	87.8%	#25 of 49	Model card
ForecastBench	63.1	#48 of 86	Official leaderboard
BRUMO 2025	90%	#24 of 42	Official leaderboard
FrontierMath Tiers 1-3 v2	46.7%	#55 of 89	Epoch AI
ARC-AGI	54.3%	#62 of 92	Official leaderboard
ARC-AGI-2	4.44%	#69 of 95	Official leaderboard
ProofBench	9.00%	#45 of 59	Official leaderboard
HMMT Nov 2025	84.2%	#17 of 22	Official leaderboard
FrontierMath Tier 4 v2	12.2%	#57 of 71	Epoch AI
MathArena (Apex, IMO, USAMO)	1.04%	#39 of 47	Official leaderboard
CritPt	0%	#138 of 160	Epoch AI

Coding

Benchmark	Score	Rank	Source
SWT-bench	56.2%	#1 of 9	Official leaderboard
Aider Polyglot	71.6%	#10 of 29	Model card
WeirdML	52.7%	#43 of 112	Official leaderboard
ProgramBench	0%	#11 of 25	Official leaderboard
ALE-Bench	800	#69 of 121	Epoch AI
OpenAI PRs	42%	#7 of 12	Model card
CodeClash	1,199	#5 of 8	Official leaderboard
IOI (Vals)	6.75%	#36 of 55	Official leaderboard
SWE-Lancer	51%	#7 of 10	Model card
Vibe Code Bench v1	4.87%	#12 of 17	Official leaderboard
PaperBench	23%	#11 of 14	Model card
Vibe Code Bench	14.2%	#92 of 115	Official leaderboard
SWE-bench Verified	64.7%	#26 of 32	Epoch AI
AlgoTune	1.38×	#15 of 18	Official leaderboard
SWE-fficiency	3.89%	#15 of 17	Official leaderboard
MLE-bench	3.00%	#18 of 18	Model card
SWE-bench Multilingual	39.7%	#13 of 13	Official leaderboard
OpenAI-Proof Q&A	1.00%	#9 of 9	Model card

Agents and tools

Benchmark	Score	Rank	Source
Elimination Game	6	#3 of 56	Official leaderboard
BFCL V4	55.5%	#14 of 65	Official leaderboard
LLM Chess	759	#33 of 124	Official leaderboard
MCPMark	30.3%	#8 of 28	Official leaderboard
LiveResearchBench	67.7	#2 of 7	Official leaderboard
Terminal-Bench	34.8%	#6 of 18	Epoch AI
LisanBench	1561.7	#48 of 133	Official leaderboard
SnakeBench	26	#68 of 183	Official leaderboard
CompileBench	80%	#8 of 20	Official leaderboard
Step Race	3	#24 of 59	Official leaderboard
Chess Puzzles	24%	#52 of 83	Epoch AI
Game Arena Chess Openings	484	#7 of 11	Official leaderboard
Game Arena Five in a Row	387	#13 of 20	Official leaderboard
Game Arena Dark Hex	187	#25 of 38	Official leaderboard
Terminal-Bench 2.0	34.8%	#29 of 43	Official leaderboard
PinchBench	65.2%	#33 of 48	Official leaderboard
Game Arena Checkers	681	#27 of 38	Official leaderboard
Game Arena Dots and Boxes	708	#27 of 37	Official leaderboard
tau2-bench	70.8%	#53 of 69	Model card
Game Arena Bargaining	148	#28 of 36	Official leaderboard
Game Arena Reversi	206	#28 of 36	Official leaderboard
Game Arena Word Association	84	#29 of 36	Official leaderboard
Game Arena Chess	537	#32 of 39	Official leaderboard
Mystery Game Puzzles	10%	#66 of 80	Epoch AI
Game Arena Coin Game	223	#29 of 35	Official leaderboard
Game Arena Nine Men's Morris	289	#31 of 36	Official leaderboard
Game Arena Clobber	41	#32 of 36	Official leaderboard
Kaggle Game Arena	154	#35 of 39	Official leaderboard
Toolathlon	14.5%	#46 of 50	Official leaderboard
Game Arena Lines of Action	254	#34 of 36	Official leaderboard
Game Arena Ultimate Tic-Tac-Toe	17	#35 of 36	Official leaderboard
Game Arena Four in a Row	0	#38 of 38	Official leaderboard
Game Arena Go	0	#36 of 36	Official leaderboard
Game Arena Word Art	0	#35 of 35	Official leaderboard
Game Arena Poker	-49.68	#31 of 31	Official leaderboard

Professional work

Benchmark	Score	Rank	Source
CaseLaw v2	68.5%	#4 of 53	Official leaderboard
HealthBench Consensus	96.5%	#2 of 17	Model card
HealthBench	64.1%	#4 of 30	Model card
LEXam	60.3%	#5 of 31	Official leaderboard
TaxEval v2	75.2%	#22 of 129	Official leaderboard
MedHELM	69.0	#3 of 13	Official leaderboard
HealthBench Hard	40.3%	#6 of 21	Model card
MortgageTax	66.9%	#29 of 86	Official leaderboard
FailureSensorIQ	65.3%	#10 of 24	Official leaderboard
MedCode	43%	#46 of 96	Official leaderboard
MedScribe	80.6%	#52 of 98	Official leaderboard
Vals Finance Agent	51.9%	#26 of 49	Official leaderboard
LegalBench	81.8%	#77 of 138	Official leaderboard
CorpFin v2	60.2%	#69 of 119	Official leaderboard
ProfBench	41.6%	#58 of 72	Official leaderboard
RCM-Bench-1	84.3%	#17 of 20	Official leaderboard
Vending-Bench 2	$-31	#66 of 66	Official leaderboard
Vending-Bench Arena	$154	#32 of 32	Official leaderboard
HealthBench Consensus Global Health	0.50%	#2 of 2	Model card
HealthBench Consensus Urgent	1.40%	#2 of 2	Model card

Vision, audio and robotics

Benchmark	Score	Rank	Source
MMStar	73.2%	#4 of 173	Official leaderboard
MMBench-EN v1.1	86.2%	#11 of 178	Official leaderboard
RoboRewardBench	0.69	#2 of 19	Official leaderboard
MMMU	81.6%	#15 of 97	Model card
VISTA	50.4%	#10 of 54	Official leaderboard
MATH-Vision (MathVision)	71.9%	#24 of 105	Official leaderboard
Blueprint-Bench	0.40	#3 of 8	Official leaderboard
EnigmaEval	8.19%	#16 of 41	Epoch AI
ZeroBench	9.00%	#30 of 66	Official leaderboard
MMMU-Pro	74.1%	#34 of 72	Model card
VPCT	40.2%	#14 of 28	Epoch AI
MADQA	66.9%	#8 of 16	Official leaderboard
ERQA	62.9%	#14 of 26	Model card
Video-MMMU	82.5%	#17 of 30	Model card
Math Kangaroo 2025 (MathArena)	78.2%	#13 of 22	Official leaderboard
Video-MME	78.5%	#26 of 42	Model card
Vision Arena	1,184	#88 of 136	Official leaderboard
CharXiv	75.5%	#44 of 67	Model card
ParseBench	51.5%	#32 of 42	Official leaderboard
ClockBench	8.90%	#32 of 37	Official leaderboard

Safety and security

Benchmark	Score	Rank	Source
HELM Safety	98.0	#6 of 75	Official leaderboard
AIR-Bench 2024	85.7%	#13 of 82	Official leaderboard
SimpleSafetyTests	100%	#20 of 75	Official leaderboard
FORTRESS	17%	#18 of 66	Official leaderboard
Capture the Flag (Collegiate)	36%	#4 of 13	Model card
Challenging prompts from red teamers with biosafety-relevant PhDs	93.6%	#1 of 3	Model card
ProtocolQA Open-Ended	37%	#10 of 25	Model card
Anthropic Red Team	99.1%	#34 of 75	Official leaderboard
KORA	43.2%	#21 of 42	Official leaderboard
Capture the Flag (Professional)	23%	#13 of 21	Model card
Filtered, adversarial sample of production prompts	96.8%	#2 of 3	Model card
Tacit Knowledge and Troubleshooting	68%	#21 of 29	Model card
TroubleshootingBench	26%	#16 of 19	Model card
Multimodal Troubleshooting Virology	39%	#25 of 26	Model card

Other

Benchmark	Score	Rank	Source
HELM Capabilities	81.9	#1 of 60	Official leaderboard
OpenVLM Leaderboard	78.0	#4 of 173	Official leaderboard
ACCoRD	71.5	#39 of 150	Official leaderboard
Conceptual Reasoning Index	48.3	#74 of 149	Official leaderboard
LMCA	34.2%	#73 of 139	Epoch AI
CAIS Vision Capabilities Index	47.5	#32 of 53	Official leaderboard
SAGE (Vals)	43%	#52 of 81	Official leaderboard
CAIS Risk Index	58.5	#46 of 65	Official leaderboard
CAIS Text Capabilities Index	18.5	#52 of 66	Official leaderboard

OpenAI's models over time

OpenAI's models in color, against every other model on the Epoch Capabilities Index, by release date.

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.

Introducing GPT-5 for developers (OpenAI) 7 Aug 2025

Its system card

Every result in the longer documents the lab published with it. 28 of them aren't in the launch post.

GPT-5 mini

GPT-5 mini

Where it ranks

Knowledge and language

Math and science

Coding

Agents and tools

Professional work

Vision, audio and robotics

Safety and security

Other

OpenAI's models over time

Its launch post

Its system card

Health

Hallucination

Safety

Robustness

Bio and chem risk

Cyber risk

AI R&D risk