Inference
EN
xAI

Grok 4.20 (Reasoning)

Grok 4.20 (Reasoning) is a model from xAI. 13 benchmarks count toward its index, in 6 categories.

  • Input: text
RankedEnough results for an index.
IndexAll categories, out of 100.55.5±5.5Median 51.0
CoverageIndex weight that has results.90%
SpeedOutput tokens per second.—Median 67/s
First tokenSeconds to the first output token.—Median 1.3 s
Input / 1MUS dollars per 1M input tokens.—Median $0.60
Output / 1MUS dollars per 1M output tokens.—Median $2.55
ContextMaximum tokens in one request.—Median 262K
EloLMArena rating and rank.—Median 1422

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.2 counted48.9135 / 218
CodingCode writing and repair.5 counted51.6127 / 209
ReasoningLogic problems and puzzles.2 counted68.625 / 227
MultimodalTasks with images and text.1 counted63.8—
KnowledgeFacts and expert knowledge.2 counted55.887 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.1 counted57.7—

Results

13 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
AIMEMath96.5% ±0.56 / 95
GPQA Diamond · Vals AIKnowledge88.6% ±1.634 / 135
ARC-AGI-2 (semi-private)Reasoning65.1%34 / 131
IOI v1Coding30.2% ±7.516 / 61
LiveCodeBenchCoding84.3% ±1.039 / 142
ARC-AGI-1 (semi-private)Reasoning89.5%37 / 126
MMLU ProKnowledge86.3% ±0.343 / 135
MMMU ProMultimodal83.5% ±0.932 / 92
Terminal-Bench 2.0 · Vals AIAgentic40.4% ±5.232 / 66
SWE-benchCoding72.2% ±2.056 / 87
Terminal-Bench 2.1 · Vals AIAgentic44.2% ±1.064 / 75
Vibe Code Bench v1.1Coding4.1% ±2.1110 / 123
Code MigrationCoding0.3% ±0.176 / 77
MedQAHealthcare94.5% ±0.216 / 95
ARC-AGI-1 (public eval)Reasoning95.5%31 / 116
ARC-AGI-2 (public eval)Reasoning63.6%34 / 122
TaxEval v2Finance74.1% ±0.941 / 143
CorpFin v2Finance63.7% ±0.940 / 131
Finance Agent v1.1Finance52.3%24 / 51
CaseLaw v2Legal54.4% ±0.337 / 53
LegalBenchLegal77.7% ±0.5107 / 149
Legal Research BenchLegal13.9% ±2.476 / 90
MedCodeHealthcare32.2% ±2.190 / 106
MortgageTaxFinance45.4% ±1.086 / 97
Tax Agent BenchFinance7.0% ±1.874 / 82
Finance Agent (v2)Finance28.5% ±0.373 / 79
MedScribeHealthcare63.4% ±2.1103 / 109
EMBFinance11.7% ±1.672 / 75

Sources

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

History

5 releases
Index56.6 → 55.5
23 Oct 202612 Nov 2026

Questions

How good is Grok 4.20 (Reasoning)?
Its index is 55.5 out of 100. That is place 98 of 283 ranked models.
What is Grok 4.20 (Reasoning) best at?
Reasoning: 68.6 out of 100, place 25 of 227.

Badge

Grok 4.20 (Reasoning) on Inference360: place 98 of 283, index 55.5
HTML
<a href="https://inference360.ai/models/xai-grok-4-20-reasoning"><img src="https://inference360.ai/badges/xai-grok-4-20-reasoning.svg" alt="Grok 4.20 (Reasoning) on Inference360: place 98 of 283, index 55.5" height="20"></a>
Markdown
[![Grok 4.20 (Reasoning) on Inference360: place 98 of 283, index 55.5](https://inference360.ai/badges/xai-grok-4-20-reasoning.svg)](https://inference360.ai/models/xai-grok-4-20-reasoning)

Compare with

MiniMax M355.7GPT-5.155.3

More from xAI

All models
Grok 4.769.1Grok 4.667.8Grok 4.563.7Grok 4.353.3Grok 4.2052.8