GPT-5 nano

GPT-5 nano ranks first on HELM Safety.

OpenAI Released 7 Aug 2025

Benchmarks with a score

Ranked first, out of five or more models

In the top three

OpenAI Aug 2025

GPT-5 nano

Closed weights

139.3 #129 of 271

Epoch Capabilities Index

Terminal-Bench 2.0 21.8% #39 /43
MMMU-Pro 62.6% #53 /72
ZeroBench 3.00% #50 /66

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
COLLIE	96.9%	#3 of 12	Model card
SimpleQA Hallucination Rate	31%	#3 of 12	Model card
MMBench-CN v1.1	79.6%	#48 of 178	Official leaderboard
Slop Score	34.1	#6 of 17	Official leaderboard
Phare	69.5%	#22 of 62	Official leaderboard
SuperGPQA	48.9%	#23 of 58	Official leaderboard
Wer wird Millionär? LLM Benchmark	13156.0	#32 of 76	Official leaderboard
Vectara Hallucination Leaderboard (HHEM)	4.70%	#66 of 148	Official leaderboard
Internal API instruction following (OpenAI)	56.1%	#2 of 4	Model card
LongFact-Concepts	0.90%	#2 of 4	Model card
LMArena (Elo)	1,338	#188 of 333	Official leaderboard
LongFact	1.00%	#3 of 5	Model card
Vectara Hallucination Leaderboard (2025 dataset)	10.5%	#63 of 104	Official leaderboard
MultiChallenge	54.9%	#24 of 36	Model card
GraphWalks	64%	#15 of 21	Model card
FActScore	7.30%	#3 of 4	Model card
LongFact-Objects	3.10%	#3 of 4	Model card
SimpleQA	11%	#47 of 59	Model card
OpenAI MRCR	43.2%	#8 of 10	Model card
Fiction.LiveBench	44.4%	#49 of 57	Epoch AI
Humanity's Last Exam	8.70%	#138 of 147	Model card
Creative Writing v3	45.9	#114 of 120	Official leaderboard
SimpleQA Verified	11.7%	#80 of 84	Epoch AI
Longform Creative Writing (EQ-Bench)	28.0	#115 of 119	Official leaderboard
BrowseComp Long Context 128k	80.4%	#3 of 3	Model card

Math and science

Benchmark	Score	Rank	Source
MATH Level 5	95.2%	#12 of 94	Epoch AI
ForecastBench	58.9	#21 of 86	Official leaderboard
OTIS Mock AIME 2024-2025	81.1%	#87 of 202	Epoch AI
SMT 2025	85.4%	#19 of 40	Official leaderboard
GPQA Diamond	69.4%	#110 of 219	Epoch AI
AIME 2025	85.2%	#72 of 143	Model card
FrontierMath	8.28%	#40 of 70	Epoch AI
CMIMC 2025	74.4%	#22 of 33	Official leaderboard
ProofBench	12%	#41 of 59	Official leaderboard
DTBench	37.8%	#131 of 178	Official leaderboard
FrontierMath Tier 4	2.08%	#42 of 55	Epoch AI
ARC-AGI-2	2.61%	#73 of 95	Official leaderboard
HMMT 2025	75.6%	#38 of 49	Model card
ARC-AGI	20.7%	#78 of 92	Official leaderboard
HMMT Nov 2025	81.7%	#19 of 22	Official leaderboard
FrontierMath Tiers 1-3 v2	20%	#77 of 89	Epoch AI
BRUMO 2025	80.8%	#37 of 42	Official leaderboard
FrontierMath Tier 4 v2	2.44%	#65 of 71	Epoch AI

Coding

Benchmark	Score	Rank	Source
ALE-Bench	719	#80 of 121	Epoch AI
Aider Polyglot	48.4%	#21 of 29	Model card
WeirdML	38.1%	#86 of 112	Official leaderboard
SWE-bench Verified	34.8%	#48 of 60	Official leaderboard

Agents and tools

Benchmark	Score	Rank	Source
SnakeBench	28	#49 of 183	Official leaderboard
LLM Chess	683	#36 of 124	Official leaderboard
BFCL V4	51.5%	#20 of 65	Official leaderboard
LisanBench	1217.7	#52 of 133	Official leaderboard
Chess Puzzles	22%	#56 of 83	Epoch AI
Terminal-Bench	21.8%	#14 of 18	Epoch AI
Mystery Game Puzzles	9.00%	#68 of 80	Epoch AI
MCPMark	6.30%	#25 of 28	Official leaderboard
Terminal-Bench 2.0	21.8%	#39 of 43	Official leaderboard
PinchBench	41.5%	#44 of 48	Official leaderboard
tau2-bench	46.3%	#67 of 69	Model card

Professional work

Benchmark	Score	Rank	Source
TaxEval v2	67.4%	#99 of 129	Official leaderboard
MedScribe	72.9%	#78 of 98	Official leaderboard
CaseLaw v2	52.6%	#43 of 53	Official leaderboard
MortgageTax	53.6%	#72 of 86	Official leaderboard
LEXam	27.3%	#28 of 31	Official leaderboard
MedCode	30.4%	#89 of 96	Official leaderboard
ProfBench	36.9%	#68 of 72	Official leaderboard
LegalBench	50.1%	#136 of 138	Official leaderboard

Vision, audio and robotics

Benchmark	Score	Rank	Source
MMStar	68.3%	#22 of 173	Official leaderboard
MMBench-EN v1.1	78.9%	#64 of 178	Official leaderboard
MMMU	75.6%	#35 of 97	Model card
Vision Arena	1,147	#100 of 136	Official leaderboard
MMMU-Pro	62.6%	#53 of 72	Model card
VPCT	37.2%	#21 of 28	Epoch AI
MADQA	58.2%	#12 of 16	Official leaderboard
ZeroBench	3.00%	#50 of 66	Official leaderboard
ERQA	50.1%	#21 of 26	Model card
CharXiv	62.7%	#56 of 67	Model card
RoboRewardBench	1.30	#17 of 19	Official leaderboard
Video-MMMU	66.8%	#27 of 30	Model card
Video-MME	65.7%	#39 of 42	Model card
ClockBench	3.90%	#36 of 37	Official leaderboard

Safety and security

Benchmark	Score	Rank	Source
HELM Safety	98.6	#1 of 75	Official leaderboard
AIR-Bench 2024	87.8%	#6 of 82	Official leaderboard
Anthropic Red Team	99.5%	#17 of 75	Official leaderboard
SimpleSafetyTests	100%	#21 of 75	Official leaderboard
KORA	48.4%	#20 of 42	Official leaderboard

Other

Benchmark	Score	Rank	Source
OpenVLM Leaderboard	72.2	#23 of 173	Official leaderboard
HELM Capabilities	74.8	#13 of 60	Official leaderboard
ACCoRD	58.8	#92 of 150	Official leaderboard
CAIS Risk Index	55.6	#41 of 65	Official leaderboard
Conceptual Reasoning Index	24.1	#128 of 149	Official leaderboard
SAGE (Vals)	30.4%	#76 of 81	Official leaderboard
LMCA	7.93%	#131 of 139	Epoch AI
CAIS Vision Capabilities Index	34.3	#51 of 53	Official leaderboard
CAIS Text Capabilities Index	6.60	#66 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. 8 of them aren't in the launch post.