OpenAI logo

o4-mini

OpenAI · Released Apr 24, 2025

ProprietaryreasoningcodingbudgetCompare this model

$1.1

Input price / MTok

#13 of 27

$4.4

Output price / MTok

#12 of 27

200K

Context window

#23 of 28

163

Tokens / sec

#7 of 21

68.1

SWE-bench Verified

#10 of 12

81.4

GPQA Diamond

#14 of 19

92.7

AIME 2025

#4 of 10

81.6

MMMU

#4 of 8

Summary

o4-mini is a proprietary language model from OpenAI, released on Apr 24, 2025.

At $1.1 input and $4.4 output per million tokens, it ranks #13 of 27 for input price, placing it mid-range on input cost.

Its context window of 200K tokens ranks #23 of 28 for context depth in this dataset.

It targets reasoning, coding, budget workloads.

Throughput is measured at 163 tokens per second (ranked #7 of 21 for throughput).

Compact reasoning model; retired from ChatGPT Feb 2026 but remains available via API.

Benchmarks

BenchmarkScoreMax%RankSource
SWE-bench Verified68.110068.1%#10/12OpenAI (self-reported at o3/o4-mini laun...
GPQA Diamond81.410081.4%#14/19OpenAI (self-reported at o3/o4-mini laun...
AIME 202592.710092.7%#4/10OpenAI (self-reported at o3/o4-mini laun...
MMMU81.610081.6%#4/8OpenAI (self-reported at o3/o4-mini laun...

Similar Models

Compare all 6 models

Models with the most shared benchmark coverage and closest scores to o4-mini.

o4-mini
GoogleGemini 2.5 Pro
OpenAIo3
OpenAIGPT-5
AnthropicClaude Sonnet 4.6
AnthropicClaude Haiku 4.5
SWE-bench Verified#10 of 12
Real-world GitHub issue resolution
GPQA Diamond#14 of 19
Graduate-level science reasoning
AIME 2025#4 of 10
Advanced math competition problems
MMMU#4 of 8
Multimodal understanding across 30 subjects

Compare o4-mini side by side

Pre-filled with 5 similar models

Open comparison

Speed source: https://artificialanalysis.ai/models/o4-mini