How good is Grok 4 Fast (Reasoning)?

Its index is 45.3 out of 100. That is place 177 of 283 ranked models.

What is Grok 4 Fast (Reasoning) best at?

Reasoning: 47.8 out of 100, place 133 of 227.

Is Grok 4 Fast (Reasoning) open source?

No. Its weights are not published.

Grok 4 Fast (Reasoning)

Grok 4 Fast (Reasoning) is a reasoning model from xAI in the Grok 4 Fast family. 18 benchmarks count toward its index, in 7 categories.

Closed weights
Reasoning
Input: text
Released 2025-09

Ranked

Index45.3±4.3Median 51.0

Coverage95%

Speed—Median 67/s

First token—Median 1.3 s

Input / 1M—Median $0.60

Output / 1M—Median $2.55

Context2MMedian 262K

Elo1400±5 #158Median 1422

Capabilities

Out of 100

Agentic3 counted42.9169 / 218

Coding4 counted44.9163 / 209

Reasoning4 counted47.8133 / 227

Multimodal1 counted42.7—

Knowledge4 counted46.6162 / 268

MultilingualNone counted——

Instruction1 counted36.2—

Math1 counted55.4—

Results

18 counted

BenchmarkResultPlace

AIMEMath91.3% ±1.025 / 95

GPQA Diamond · Vals AIKnowledge85.4% ±1.848 / 135

Artificial Analysis GPQA DiamondKnowledge84.7%78 / 199

Critical Physics TasksReasoning2.9%98 / 214

Artificial Analysis Long Context ReasoningReasoning73.7%100 / 217

LiveCodeBenchCoding79.0% ±1.174 / 142

Artificial Analysis Humanity's Last ExamKnowledge19.1%116 / 222

IOI v1Coding11.5% ±4.433 / 61

Artificial Analysis IFBenchInstruction50.5%79 / 146

τ²-Bench Tool-Agent-User EvaluationAgentic65.8%81 / 149

ARC-AGI-1 (semi-private)Reasoning48.5%75 / 126

ARC-AGI-2 (semi-private)Reasoning5.3%79 / 131

MMMU ProMultimodal72.8% ±1.161 / 92

Artificial Analysis LiveCodeBenchCoding83.2%8 / 12

MMLU ProKnowledge79.7% ±0.493 / 135

Terminal-Bench 2.0 · Vals AIAgentic29.2% ±4.846 / 66

Artificial Analysis Omniscience AccuracyKnowledge22.8%117 / 164

Terminal-Bench 1.0Agentic27.5% ±5.035 / 47

Artificial Analysis MMMU-ProMultimodal61.8%89 / 113

Artificial Analysis MMLU-ProKnowledge85.0%14 / 15

SWE-benchCoding45.4% ±2.283 / 87

Vibe Code Bench v1.1Coding0.0%120 / 123

TaxEval v2Finance75.7% ±0.812 / 143

CorpFin v2Finance66.9% ±0.915 / 131

CaseLaw v2Legal65.7% ±0.410 / 53

MedQAHealthcare92.1% ±0.235 / 95

MGSMMultilingual90.9% ±1.831 / 74

MedScribeHealthcare81.6% ±2.153 / 109

LegalBenchLegal80.6% ±0.584 / 149

ARC-AGI-2 (public eval)Reasoning4.6%71 / 122

ARC-AGI-1 (public eval)Reasoning58.1%71 / 116

Finance Agent v1.1Finance46.1% ±2.833 / 51

MedCodeHealthcare37.4% ±1.975 / 106

MortgageTaxFinance42.1% ±0.990 / 97

Artificial Analysis AIME 2025Math89.7%—

Sources

BenchLMCC BY-NC 4.0 Vals AINo licence stated ARC Prize FoundationTerms restrict reuse

History

5 releases

Index46.1 → 45.3

23 Oct 202612 Nov 2026

Questions

How good is Grok 4 Fast (Reasoning)?: Its index is 45.3 out of 100. That is place 177 of 283 ranked models.
What is Grok 4 Fast (Reasoning) best at?: Reasoning: 47.8 out of 100, place 133 of 227.
What is the context window of Grok 4 Fast (Reasoning)?: 2M tokens.
Is Grok 4 Fast (Reasoning) open source?: No. Its weights are not published.
When was Grok 4 Fast (Reasoning) released?: September 2025.

Badge

HTML: <a href="https://inference360.ai/models/grok-4-fast-reasoning"><img src="https://inference360.ai/badges/grok-4-fast-reasoning.svg" alt="Grok 4 Fast (Reasoning) on Inference360: place 177 of 283, index 45.3" height="20"></a>
Markdown: [![Grok 4 Fast (Reasoning) on Inference360: place 177 of 283, index 45.3](https://inference360.ai/badges/grok-4-fast-reasoning.svg)](https://inference360.ai/models/grok-4-fast-reasoning)

Compare with

Qwen3 Max45.0 GLM-4.545.7 o4-mini44.7 Mercury 2.546.2 Grok 4.1 Fast (Reasoning)47.3

More from xAI

All models

Grok Code Fast 141.2 Grok 4.1Unranked Grok 4.1 Fast (Reasoning)47.3 Grok 4.1 Fast33.3 Grok 450.0