Note: Overall leaderboard rankings may not reflect true model quality — individual benchmarks give a clearer picture. ARC-Challenge MMLU GPQA GSM8K Artificial Analysis Intelligence Index v4.1
← Back to leaderboard

OpenAI: o3

4 benchmarks
CompanyOpenAI
Country of originUSA

Leaderboard position

#218
Overall rank (of 1121 models)
1834
Elo rating
4
Benchmarks
Top 19.4%
Percentile
#25
In OpenAI (of 74 models)
Above: Anke Tang: clip-vit-base-patch16_cifar10 #217 Below: H company: holo2-8b #219

Benchmark highlights

Best benchmark
SuperGPQA 62.79
Worst benchmark
Artificial Analysis Intelligence Index v4.1 (Maximum Reasoning) 30

Nearest competitor from another company

The model from another company with the most similar benchmark scores.

appeared as nearest in 2 of 4 benchmarks

SpaceXAI: Grok 4
2/4 50.0%
Model Votes % of benchmarks
SpaceXAI: Grok 4 2 50.0%
inclusionAI: ring-2.6-1t 1 25.0%
Anthropic: claude-4.5-haiku 1 25.0%

Benchmarks

SuperGPQA 62.79 Dr. Bench 61.76 Artificial Analysis Intelligence Index v4.0 (Maximum Reasoning) 38.4 Artificial Analysis Intelligence Index v4.1 (Maximum Reasoning) 30

Predicted scores

Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.

Benchmark Predicted Confidence Based on
MATH-500 ~99.2 60% Qwen: Qwen3 8B, OpenBMB: minicpm5-1b
AIME 2025 ~98.24 60% Moonshot AI: Kimi K2 Thinking, OpenAI: gpt-oss-20b, Qwen: Qwen3 8B
AIME 2026 ~92.44 60% Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B, Qwen: qwen3.5-397b-a17b
HMMT Feb 2026 ~84.02 60% Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B, Qwen: qwen3.5-397b-a17b
GPQA Diamond ~79.63 60% Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B, Qwen: qwen3.6-27b
MMLU-Pro ~85.26 60% Qwen: qwen3.5-122b-a10b, Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B
Humanity's Last Exam (no tools) ~29.07 60% Qwen: Qwen3.5 27B, Moonshot AI: Kimi K2 Thinking, Qwen: qwen3.6-27b
IFEval ~95.66 60% Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B, OpenAI: o3-mini
IFBench ~69.55 60% Qwen: Qwen3.5 27B, Meta Llama: Llama 3.3 70B Instruct, OpenBMB: minicpm5-1b
HumanEval ~95.9 60% OpenAI: gpt-4o
MBPP ~93.7 60% OpenAI: gpt-4o
SWE-bench Verified ~69.92 60% Qwen: Qwen3.5 27B, Qwen: qwen3.6-27b, OpenAI: gpt-oss-120b
SWE-bench Pro ~46.8 60% Qwen: Qwen3.5 27B, Qwen: qwen3.6-27b, OpenAI: gpt-oss-120b
SciCode ~42.22 60% OpenAI: o3-mini, Meta Llama: Llama 3.3 70B Instruct
AA-LCR ~62.24 60% Meta Llama: Llama 3.3 70B Instruct
Humanity's Last Exam (with tools) ~39.13 60% Moonshot AI: Kimi K2 Thinking, OpenAI: gpt-oss-120b, OpenAI: gpt-oss-20b
GSM8K ~99.6 60% OpenAI: gpt-oss-20b, DeepSeek: deepseek-v3, Meta Llama: Llama 3.1 405B Instruct
-(CoderPPL) ~-1.69 60% Qwen: Qwen3.5 9B, OpenAI: gpt-oss-20b, OpenBMB: minicpm5-1b
LiveCodeBench v6 ~78.31 60% Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B, Qwen: qwen3.6-27b
MMMU-Pro ~71.34 60% Qwen: Qwen3.5 27B, Qwen: qwen3.6-27b
MATH ~97.9 60% OpenAI: o3-mini
BBH ~87.07 60% OpenBMB: minicpm5-1b
ARC-Easy ~94.8 60% Qwen: Qwen3.5 9B, OpenBMB: minicpm5-1b, Qwen: Qwen3.5 0.8B
ARC-Challenge ~97.3 60% Qwen: Qwen3.5 9B, Meta Llama: Llama 3.1 405B Instruct, OpenBMB: minicpm5-1b
HellaSwag ~95.4 60% Qwen: Qwen3.5 9B, OpenBMB: minicpm5-1b, Qwen: Qwen3.5 0.8B