Models/Grok-3 mini

xAI: Grok-3 mini

Watch this model Compare Find source

grok-3-mini

Grok-3 mini is xAI’s proprietary AI model, released Jun 24, 2025. It scores 140 on the Epoch Capabilities Index.

Context

—

Input

— / 1M

Output

— / 1M

Released

Jun 24, 2025

Modalities

Benchmarks

Scores on standardized evaluations. Higher is better — and rank shows where Grok-3 mini lands among all models tracked on Model Beat.

23.0

Intelligence Index

Epoch AI

23rd percentile of tracked models

Reasoning

5 evals

ARC-AGI16.5%

Abstract visual reasoning

ARC-AGI-20.4%

Harder abstract reasoning

GPQA Diamond76.3%

Graduate-level scientific reasoning

SimpleQA Verified21.1%

Factual accuracy & hallucination

WeirdML42.6%

Novel ML problem-solving

Math

3 evals

AIME 2024/202577.8%

Olympiad-qualifier math

FrontierMath5.9%

Research-level math problems

MATH Level 590.9%

Hardest competition math

Compare Grok-3 mini with

Grok-3 mini vs Claude Opus 5.5 Grok-3 mini vs GPT-6 Astra Grok-3 mini vs Kimi K3 Grok-3 mini vs Gemini 3.7 Flash Grok-3 mini vs Grok 4.6

Frequently asked questions

What is Grok-3 mini?

Grok-3 mini is a Grok-family AI model, from xAI, released Jun 24, 2025.

Who created Grok-3 mini?

Grok-3 mini is an AI model developed by xAI, part of the Grok family, released Jun 24, 2025.

How does Grok-3 mini perform on benchmarks?

On standardized evaluations tracked by Epoch AI, Grok-3 mini scores 90.9% on MATH Level 5, 42.6% on WeirdML, 77.8% on AIME 2024/2025.

Is Grok-3 mini open source?

Grok-3 mini is a proprietary model (API access), available through its provider's API rather than as a downloadable open-weight model.

Benchmark data from Epoch AI (CC BY) and Artificial Analysis; pricing, specs & descriptions from OpenRouter. Data updated Jun 23, 2026.