o3-pro

o3-pro ranks first on Fiction.LiveBench.

OpenAI Released June 10th, 2025 Compare with others

Benchmarks with a score

Ranked first, out of five or more models

In the top three

OpenAI Jun 2025

Closed weights

147.5 #81 of 256

Epoch Capabilities Index

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.

Benchmark	Score	Rank	Source
Fiction.LiveBench	97.2%	#1 of 57	Fiction.live leaderboard, via Epoch AI
MultiNRC	49%	#10 of 38	Official leaderboard
LLM Confabulation Leaderboard	14.2%	#18 of 58	Official leaderboard
MultiChallenge	62.4%	#14 of 42	Official leaderboard
Vectara Hallucination Leaderboard (2025 dataset)	23.3%	#102 of 104	Official leaderboard

Benchmark	Score	Rank	Source
IneqMath	46%	#2 of 34	Official leaderboard
AIME 2024	93%	#8 of 88	Lab model card
GPQA Diamond	84%	#88 of 302	Lab model card
DTBench	86.9%	#67 of 180	Official leaderboard
ARC-AGI	59.3%	#62 of 95	Official leaderboard
ARC-AGI-2	4.86%	#69 of 98	Official leaderboard

Benchmark	Score	Rank	Source
Aider Polyglot	84.9%	#2 of 46	Official leaderboard
Codeforces rating	2,748	#6 of 47	Lab model card
WeirdML	58.2%	#36 of 112	Official leaderboard

Benchmark	Score		Rank	Source
MCP Atlas	44.5%		#36 of 37	Official leaderboard

Benchmark	Score		Rank	Source
PRBench	49.7%		#12 of 48	Official leaderboard

Benchmark	Score		Rank	Source
VISTA	51.6%		#6 of 54	Official leaderboard
ClockBench	14.4%		#24 of 37	Official leaderboard

Benchmark	Score	Rank	Source
ACCoRD	73.8	#32 of 152	Official leaderboard
TutorBench	54.6%	#5 of 20	Official leaderboard
Conceptual Reasoning Index	53.5	#58 of 151	Official leaderboard
LMCA	38.5%	#62 of 147	Official leaderboard

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

Source: Epoch Capabilities Index, Epoch AI (CC BY 4.0)

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