o4-mini - In-Depth Overview

OpenAI · o-series

Current

Latest in family

Model ID: o4-mini

Get all the details on o4-mini, an AI model from OpenAI. This page covers its token limits, pricing structure, key capabilities such as multimodal_input, long_context, function_calling, available API code samples, and performance strengths.

Key Metrics

Input Limit

200K tokens

Output Limit

100K tokens

Input Cost

$1.10/1M

Output Cost

$4.40/1M

Sample API Code

from openai import OpenAI;client = OpenAI();response = client.chat.completions.create(model="o4-mini",messages=[{"role": "system", "content": "You are a helpful assistant."},{"role": "user", "content": "What is the capital of France?"}],max_tokens=100);print(response.choices[0].message.content)

Required Libraries

openai

Benchmarks

Benchmark	Score	Source	Notes
lmarena text	1345	LM Arena Leaderboard	Score for o4-mini-2025-04-16 snapshot.
lmarena webdev	1095	LM Arena Leaderboard	Score for o4-mini-2025-04-16 snapshot.
lmarena vision	1270	LM Arena Leaderboard	Score for o4-mini-2025-04-16 snapshot.
gpqa	81.4%	Vellum LLM Leaderboard	GPQA Diamond benchmark.
aime 2024	93.4%	Vellum LLM Leaderboard	High School Math (AIME 2024) benchmark.
swe bench	68.1%	Vellum LLM Leaderboard	Agentic Coding (SWE Bench) benchmark.
humanitys last exam	14.28	Vellum LLM Leaderboard	Best Overall (Humanity's Last Exam) benchmark.
grind	50%	Vellum LLM Leaderboard	Adaptive Reasoning (GRIND) benchmark.
alder polyglot	68.9%	Vellum LLM Leaderboard	Alder Polyglot benchmark.
livebench global average	78.72	LiveBench Leaderboard	Score for o4-Mini High.
livebench reasoning average	88.11	LiveBench Leaderboard	Score for o4-Mini High.
livebench coding average	79.98	LiveBench Leaderboard	Score for o4-Mini High.
livebench mathematics average	84.90	LiveBench Leaderboard	Score for o4-Mini High.
livebench data analysis average	68.33	LiveBench Leaderboard	Score for o4-Mini High.
livebench language average	66.05	LiveBench Leaderboard	Score for o4-Mini High.
livebench if average	84.96	LiveBench Leaderboard	Score for o4-Mini High.

Notes

o4-mini is OpenAI's latest small o-series model, optimized for fast, effective reasoning with exceptionally efficient performance in coding and visual tasks. It supports a 200,000 context window and 100,000 max output tokens. It also supports reasoning tokens.