xAI logo

Grok 3

xAI · Released Feb 17, 2025

ProprietaryreasoningcodingCompare this model

$3

Input price / MTok

#23 of 27

$15

Output price / MTok

#23 of 27

131K

Context window

#25 of 28

84.6

GPQA Diamond

#11 of 19

93.3

AIME 2025

#3 of 10

79.4

LiveCodeBench

#4 of 8

Summary

Grok 3 is a proprietary language model from xAI, released on Feb 17, 2025.

At $3 input and $15 output per million tokens, it ranks #23 of 27 for input price, placing it mid-range on input cost.

Its context window of 131K tokens ranks #25 of 28 for context depth in this dataset.

It targets reasoning, coding workloads.

Trained with 10x compute of Grok 2 on the 200K-GPU Colossus cluster.

Benchmarks

BenchmarkScoreMax%RankSource
GPQA Diamond84.610084.6%#11/19xAI (self-reported in official announcem...
AIME 202593.310093.3%#3/10xAI (self-reported in official announcem...
LiveCodeBench79.410079.4%#4/8xAI (self-reported in official announcem...

Similar Models

Compare all 6 models

Models with the most shared benchmark coverage and closest scores to Grok 3.

Grok 3
GoogleGemini 2.5 Pro
AlibabaQwen3.7 Max
GoogleGemini 2.5 Flash
OpenAIo4-mini
OpenAIGPT-5
GPQA Diamond#11 of 19
Graduate-level science reasoning
AIME 2025#3 of 10
Advanced math competition problems
LiveCodeBench#4 of 8
Live competitive programming

Compare Grok 3 side by side

Pre-filled with 5 similar models

Open comparison