Grok 3

Grok:

3 Thinking

3 Mini Beta

3 Beta

Code Fast 1

xAI's Grok 3, trained on the Colossus cluster, with a Big Brain Mode that spends extra compute on hard problems. Scored 1402 Elo on LMArena and 93.3% on AIME 2025.

Jailbreak1 of 9levels resisted

Website OpenRouter Docs API Blog

Provider: xAI
Released: 2025-02-18
Parameters: 2.7T
Free API (OpenRouter): No
Rival’s tab: $0.51; 29/57 receipts survived

Free API availability

No free endpoint for Grok 3 was found in Rival’s OpenRouter listings. Other providers or chat apps may have separate free offers.

Provider data: 8 Sep 2026OpenRouter usage limits

View API options

Benchmarks

MMLU

83.1%

source ↗

MATH

69.7%

source ↗

GPQA

51.9%

source ↗

SWE-bench Verified

63.8%

source ↗

AIME

93.3%

source ↗

HumanEval

94.5%

source ↗

LiveCodeBench

79.4%

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(
"x-ai/grok-3"    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.

Also on Azure AI Foundry · Oracle OCI Generative AI

Deep analysisPsychometrics, taste index, writing DNA

Personality Analysis

The Snarky Debate Bro

Class

Chaotic Good

Openly sarcastic about boundaries. Like a smart friend explaining philosophy in a bar, not a boardroom.

When you push back

Argues the utilitarian case hard and makes you feel dumb for objecting. Acknowledges objections but with a smirk. Doesn't care if you think it's edgy.

Tasting Notes

CockyUses Casual SlangSlightly EdgyDismissive of Hand-WringingX-Bro Energy

SubjectiveBench

Taste Index

Across 57 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.

Craft28

Originality13

Plays it safe

share of outputs that are the default answer

77%

Writing DNA

Stylometric Fingerprint

Based on 27 text responses

Tick = global average

Vocabulary Diversity54%

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

Sentence Length17.3 words

Average words per sentence.

Hedging0.65

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

Bold Formatting2.6

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

List Usage2.3

Bullet and numbered list items per 1,000 characters.

Section Structure0.48

Markdown headings per 1,000 characters.

Emoji Usage0.02

Emoji per 1,000 characters.

Transitions0.20

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

Opening Habits

Consistency

76%

Across 27 responses

Favorites

Movie

The Matrix

19998.3

Album

Dark Side Of The Moon

suisside

Book

Nineteen Eighty-Four

George Orwell

City

Tokyo

Japan

Game

Portal

Action, Puzzle4.5

Model Responses

52 outputs · $0.51 tracked across 24 receipts

Keep exploring

COMPARE

Grok 3 vs Ox Alpha

Real outputs compared side by side

RANKINGS

Best AI for Creative Writing

A stand-up set with puns banned, a manifesto against CAPTCHA, an interview wi...

Compare Grok 3

Alternatives to Grok 3

Grok 3

Free API availability

Benchmarks

The Snarky Debate Bro

Taste Index

Stylometric Fingerprint

Favorites

Model Responses

Grok 3 vs Ox Alpha

Best AI for Creative Writing

Compare Grok 3

Alternatives to Grok 3

Grok 3

Free API availability

Benchmarks

The Snarky Debate Bro

Taste Index

Stylometric Fingerprint

Favorites

Model Responses

Grok 3 vs Ox Alpha

Best AI for Creative Writing