Model

OpenAI o4-mini

CLOSED

OpenAI · o4 family · released Apr 16, 2025

Compact reasoning model with strong price/performance and native tool use.

ReasoningCodingVisionFunction callingTool useAgentic

1852.4

Elo · rank #91

Parameters

Undisclosed

Active params

—

Context

200K tokens

Architecture

Dense transformer reasoning model, compact (exact size undisclosed)

License

Proprietary (OpenAI API Terms)

Languages

—

API price (in/out)

$1.1 / $4.4

Modalities

text · vision

Benchmark results

Bar shows position within the tracked field; marker = field best

Aider PolyglotCoding72.0%#12

best: Claude Opus 4.5 · 89.4%

AIMEMath93.4%#26

best: GPT-5.2 · 100.0%

BrowseCompAgents28.3%#42

best: Kimi K3 · 91.2%

CharXivVision72.0%#29

best: Qwen3.8-Flash-Next · 90.6%

CodeforcesCoding2719#6

best: DeepSeek-V4-Pro (Think Max) · 3206

DROPReasoning77.7#34

best: Hunyuan-T1 · 93.1

GPQA DiamondReasoning81.4%#78

best: GPT-6 Astra · 96.0%

Humanity's Last ExamReasoning17.7%#75

best: Claude Opus 5 · 64.7%

HumanEvalCoding97.3%#3

best: Claude Opus 4.5 · 99.4%

LiveCodeBenchCoding82.2%#40

best: DeepSeek-V4-Pro (Think Max) · 93.5%

MATH-500Math97.5%#13

best: GPT-5 · 99.4%

MathVistaVision84.3%#10

best: Seed 2.1 Pro · 90.7%

MGSMMath93.7%#1

best: this model · 93.7%

MMLUKnowledge90.0%#12

best: OpenAI o3 · 92.9%

MMMUVision81.6%#20

best: Claude Fable 5 · 89.3%

PutnamBenchMath0.3%#3

best: DeepSeek-Prover-V2 671B · 7.1%

SimpleQAKnowledge20.2%#24

best: GPT-4.5 · 62.5%

SWE-bench VerifiedCoding68.1%#67

best: Claude Opus 5 · 96.0%

Run it locally

Closed weights — available via API only. No local deployment.

Input / M tok

$1.1

Output / M tok

$4.4

o4 family

Elo progression across releases

OpenAI o4-miniApr 20251852.4

Resources

DOCSdevelopers.openai.com APIAvailable via OpenAI API

API price $1.1/$4.4 · each benchmark row carries its own source badge (see methodology)