Inference
EN
xAI

Grok 4.5

Grok 4.5 is a reasoning model from xAI. 47 benchmarks count toward its index, in 7 categories.

  • Closed weights
  • Reasoning
  • Input: text · image · file
  • Released 2026-07
RankedEnough results for an index.
IndexAll categories, out of 100.63.7±2.9Median 51.0
CoverageIndex weight that has results.95%
SpeedOutput tokens per second.53/sMedian 67/s
First tokenSeconds to the first output token.1.8 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$2.00Median $0.60
Output / 1MUS dollars per 1M output tokens.$6.00Median $2.55
ContextMaximum tokens in one request.500KMedian 262K
EloLMArena rating and rank.1448±4 #59Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
x-ai/grok-4.5
Max outputMaximum output tokens per request.
450K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.30
WeightsLink to the published weights.
—
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
2 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
OpenRouterRouter$2.00$6.00500K
xAIMaker$2.00$6.00500K

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.12 counted61.662 / 218
CodingCode writing and repair.14 counted65.837 / 209
ReasoningLogic problems and puzzles.8 counted63.352 / 227
MultimodalTasks with images and text.1 counted47.7—
KnowledgeFacts and expert knowledge.6 counted65.840 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted71.3—
MathMath problems.5 counted61.841 / 168

Results

47 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
SkillsBenchCoding66.0% ±4.62 / 35
Artificial Analysis GPQA DiamondKnowledge93.1%12 / 199
VulcanBench v3Coding89.9%1 / 14
GPQA diamond · Epoch AIKnowledge93.4% ±1.418 / 217
GPQA Diamond · Vals AIKnowledge92.9% ±1.312 / 135
Artificial Analysis Omniscience AccuracyKnowledge51.6%16 / 164
MMLU ProKnowledge89.2% ±0.314 / 135
LiveCodeBenchCoding87.4% ±1.015 / 142
LiveBench ResearchAgentic85.9%6 / 52
LiveBench Coding 2.0Coding94.3%7 / 52
τ²-bench BankingAgentic47.9%4 / 28
SWE-benchCoding86.6% ±1.513 / 87
Artificial Analysis Humanity's Last ExamKnowledge42.7%36 / 222
OTIS Mock AIME 2024-2025Math97.8% ±1.333 / 200
Artificial Analysis MMMU-ProMultimodal80.4%21 / 113
SWE-bench ProCoding64.7%17 / 83
Critical Physics TasksReasoning15.4%48 / 214
Artificial Analysis Long Context ReasoningReasoning79.3%52 / 217
Artificial Analysis SciCodeCoding55.0%29 / 114
Terminal-Bench 2.1 · BenchLMAgentic83.3%24 / 86
GDPval-AA normalizedAgentic44.2%41 / 133
LiveBench LanguageKnowledge82.8%21 / 66
LiveBench Instruction FollowingInstruction71.5%22 / 66
LiveBench ReasoningReasoning87.2%29 / 80
LiveBench DesignAgentic78.1%19 / 52
SimpleQA VerifiedKnowledge48.3% ±1.634 / 91
Agent Arena command recoveryAgentic5.119 / 50
ARC-AGI-2 (semi-private)Reasoning52.6%50 / 131
cursorBench32Coding66.7%7 / 18
LiveBench Agentic CodingAgentic56.5%26 / 66
ARC-AGI-1 (semi-private)Reasoning85.7%50 / 126
Vibe Code Bench v1.1Coding69.0% ±4.551 / 123
Terminal-Bench 2.1 · Vals AIAgentic67.8% ±4.432 / 75
Code MigrationCoding36.6% ±4.134 / 77
LiveBench MathematicsMath90.8%30 / 66
Artificial Analysis Tau3-BankingAgentic42.1%11 / 22
Chess PuzzlesReasoning35.0% ±4.846 / 90
FrontierMath-Tiers-1-3-v2-PrivateMath57.2% ±2.947 / 90
Terminal-Bench 3.0Agentic15.7%7 / 13
LiveBench Data AnalysisReasoning73.0%46 / 80
FrontierMath-Tier-4-v2-PrivateMath24.4% ±6.846 / 72
LiveBench Browser UseAgentic44.0%32 / 49
LiveBench Long AgenticAgentic27.9%34 / 51
ProofBench v1.1Math31.0% ±4.638 / 54
MirrorCodeCoding32.2% ±10.728 / 39
IOICoding40.6% ±4.034 / 45
Artificial Analysis Terminal-Bench v2.1Agentic81.6%16 / 21
ARC-AGI-3 (semi-private)Reasoning0.3%20 / 26
Terminal-Bench 4.0Agentic8.6% ±1.347 / 61
DeepSWEAgentic53.0%50 / 59
PostTrainBench v1.1Coding23.4%13 / 15
MMMU ProMultimodal61.8% ±1.282 / 92
Terminal-Bench 4.0.0Agentic12.4% ±1.321 / 23
Terminal-Bench 4.1.0Agentic12.4% ±1.321 / 23
Furniture AssemblyReasoning22.5% ±5.434 / 37
LiveBench CodingCoding68.6%62 / 66
CorpFin v2Finance67.4% ±0.912 / 131
LegalBenchLegal86.0% ±0.417 / 149
MedScribeHealthcare86.9% ±1.919 / 109
ARC-AGI-1 (public eval)Reasoning94.8%35 / 116
ARC-AGI-2 (public eval)Reasoning58.2%43 / 122
MedCodeHealthcare43.3% ±2.346 / 106
CyberBench v1.1Other71.0% ±5.821 / 48
Legal Research BenchLegal38.0% ±3.443 / 90
TaxEval v2Finance71.7% ±0.974 / 143
ProgramBenchCoding0.0%34 / 65
Tax Agent BenchFinance24.9% ±3.145 / 82
MortgageTaxFinance61.8% ±1.059 / 97
Finance Agent (v2)Finance48.3% ±0.550 / 79
EMBFinance52.9% ±3.152 / 75
SRE BenchOther0.8% ±0.528 / 36
Time Horizon Index: KSPOther6.3%11 / 13
Vals Web Search IndexOther35.5% ±1.912 / 12
Artificial Analysis EnterpriseOps-GymAgentic40.8%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseEpoch AICC BYLiveBenchNo licence statedVals AINo licence statedARC Prize FoundationTerms restrict reuseSierra ResearchMITTerminal-BenchNo licence statedLMArenaCC BY 4.0models.devMIT

History

5 releases
Index64.9 → 63.7
23 Oct 202612 Nov 2026
Output / 1M$6.00 → $6.00
23 Oct 202612 Nov 2026

Questions

How much does Grok 4.5 cost?
$2.00 per 1M input tokens, $6.00 per 1M output tokens.
How good is Grok 4.5?
Its index is 63.7 out of 100. That is place 55 of 283 ranked models.
What is Grok 4.5 best at?
Knowledge: 65.8 out of 100, place 40 of 268.
How fast is Grok 4.5?
53 output tokens per second, 1.8 s to the first token.
What is the context window of Grok 4.5?
500K tokens.
Is Grok 4.5 open source?
No. Its weights are not published.
When was Grok 4.5 released?
July 2026.

Badge

Grok 4.5 on Inference360: place 55 of 283, index 63.7
HTML
<a href="https://inference360.ai/models/grok-4-5"><img src="https://inference360.ai/badges/grok-4-5.svg" alt="Grok 4.5 on Inference360: place 55 of 283, index 63.7" height="20"></a>
Markdown
[![Grok 4.5 on Inference360: place 55 of 283, index 63.7](https://inference360.ai/badges/grok-4-5.svg)](https://inference360.ai/models/grok-4-5)

Compare with

Gemini 3.6 Flash63.7dots3-note Preview63.4Qwen3.8 Max Preview64.2Grok 4.667.8Grok Build 0.152.7

Lower price

dots3-note Preview63.4 · FreeGLM-5.3-Flash63.1 · $0.078/1MLing 3.1 Flash67.2 · $0.135/1M

More from xAI

All models
Grok LatestUnrankedGrok 4.667.8Grok Build 0.152.7Grok 4.353.3Grok 4.769.1