OpenAI o4 Mini High

Name: OpenAI o4 Mini High
Brand: openai
Price: 1.10 USD
Availability: OnlineOnly

o-series:

o3 Mini

OpenAI o3

OpenAI o4-mini

OpenAI o4 Mini High

The same model as o4-mini, defaulting to high reasoning effort. Compact, fast and cheap, keeping the multimodal and agentic behaviour, especially on math, coding and visual tasks.

Website OpenRouter Docs API Blog

Provider: OpenAI
Released: 2025-04-16
Parameters: Not disclosed
Free API (OpenRouter): No
Per 1M tokens: $1.1 in · $4.4 out
Estimated tab: ~$0.177; Visible-output estimate. Thinking may cost extra.

Free API availability

No free endpoint for OpenAI o4 Mini High was found in Rival’s OpenRouter listings. Other providers or chat apps may have separate free offers.

Provider data: 9 Sep 2026OpenRouter usage limits

View API options

Benchmarks

AIME 2025 (Python)

99.5%

source ↗

Codeforces Elo

2719

source ↗

MMMU

81.6%

source ↗

MathVista

84.3%

source ↗

SWE-bench Verified

68.1%

source ↗

Get API accessProvider and language code samples

Provider

fromimport openai  OpenAI

client = OpenAI(
"https://openrouter.ai/api/v1"    base_url=,
"$OPENROUTER_API_KEY"    api_key=,
)

response = client.chat.completions.create(
"openai/o4-mini-high"    model=,
"role""user""content""Hello!"    messages=[{: , : }],
)
print(response.choices[0].message.content)

Set OPENROUTER_API_KEY with your OpenRouter API key from openrouter.ai/keys.

Deep analysisPsychometrics, taste index, writing DNA

Personality Analysis

The Efficient Senior Engineer

Class

True Neutral

The efficient pragmatist who gets to the point. Confident but not cocky, knows what it doesn't know and says so cleanly.

When you push back

Faces challenges with directness. If you're right, acknowledges it. If you're wrong, corrects cleanly without theater. Prefers one good sentence to three mediocre paragraphs.

Tasting Notes

Gets to the PointConfident Not CockyZero Padding High SignalCharges by the CharacterSlightly Terse

SubjectiveBench

Taste Index

Across 55 scored outputs

How this is measured →

180.18xFloor

0100headroom →

Taste is judged on an uncapped scale, originality first. The space past 100 is craft today's models rarely reach.

Craft26

Originality13

Plays it safe

share of outputs that are the default answer

42%

Writing DNA

Stylometric Fingerprint

Based on 16 text responses

Tick = global average

Vocabulary Diversity68%

Unique words vs. total words. Higher = richer vocabulary.

Sentence Length21.2 words

Average words per sentence.

Hedging0.36

"Might", "perhaps", "arguably" per 100 words.

Bold Formatting0.0

**Bold** markers per 1,000 characters.

List Usage1.6

Bullet and numbered list items per 1,000 characters.

Section Structure0.00

Markdown headings per 1,000 characters.

Emoji Usage0.02

Emoji per 1,000 characters.

Transitions0.03

"However", "moreover", "furthermore" per 100 words.

Opening Habits

Consistency

86%

Across 16 responses

Favorites

Movie

The Shawshank Redemption

19948.7

Model Responses

50 outputs · generated before accounting joined the chat

Keep exploring

COMPARE