Inference
EN
xAI

Grok 4.3

Grok 4.3 is a reasoning model from xAI. 27 benchmarks count toward its index, in 7 categories.

  • Closed weights
  • Reasoning
  • Input: text · image · file
  • Released 2026-04
RankedEnough results for an index.
IndexAll categories, out of 100.53.3±3.6Median 51.0
CoverageIndex weight that has results.95%
SpeedOutput tokens per second.61/sMedian 67/s
First tokenSeconds to the first output token.1.3 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$1.25batch $1.00Median $0.60
Output / 1MUS dollars per 1M output tokens.$2.50batch $2.00 Batch price per 1M output tokens.Median $2.55
ContextMaximum tokens in one request.1MMedian 262K
EloLMArena rating and rank.1397±4 #162Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
x-ai/grok-4.3
Max outputMaximum output tokens per request.
900K
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.
$0.20
WeightsLink to the published weights.
—
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
4 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
Amazon BedrockCloud$1.25$2.501M
Google Vertex AICloud$1.25$2.50200K
OpenRouterRouter$1.25$2.501M
Oracle CloudCloud$1.25$2.501M

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.7 counted50.6123 / 218
CodingCode writing and repair.8 counted53.2115 / 209
ReasoningLogic problems and puzzles.3 counted49.1121 / 227
MultimodalTasks with images and text.1 counted56.8—
KnowledgeFacts and expert knowledge.5 counted57.081 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.2 counted59.315 / 63
MathMath problems.1 counted57.3—

Results

27 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
Artificial Analysis IFBenchInstruction81.3%3 / 146
τ²-Bench Tool-Agent-User EvaluationAgentic97.7%9 / 149
Instruction Following BenchmarkInstruction81.3%8 / 57
GPQA Diamond · Vals AIKnowledge91.4% ±1.422 / 135
Artificial Analysis GPQA DiamondKnowledge90.1%35 / 199
Graduate-Level Google-Proof Q&AKnowledge90.1%29 / 141
Artificial Analysis Humanity's Last ExamKnowledge37.2%58 / 222
LiveCodeBenchCoding84.5% ±1.038 / 142
Artificial Analysis MMMU-ProMultimodal78.1%38 / 113
Scientific Code BenchmarkCoding47.3%12 / 35
Critical Physics TasksReasoning8.0%75 / 214
MMMU ProMultimodal83.1% ±0.934 / 92
MMLU ProKnowledge85.8% ±0.353 / 135
Artificial Analysis Omniscience AccuracyKnowledge34.6%69 / 164
Terminal-Bench 2.0 · Vals AIAgentic43.4% ±4.428 / 66
IOI v1Coding15.3% ±6.329 / 61
GDPval-AA normalizedAgentic29.2%64 / 133
Massive Multi-discipline Multimodal Understanding ProMultimodal78.1%26 / 53
Artificial Analysis SciCodeCoding48.3%58 / 114
Humanity's Last ExamKnowledge35.0%40 / 77
Gert Labs Composite Game BenchmarkAgentic43.9%30 / 52
Artificial Analysis Long Context ReasoningReasoning64.3%136 / 217
SWE-benchCoding71.4% ±2.057 / 87
APEX-Agents-AAAgentic17.0%20 / 30
Vibe Code Bench v1.1Coding19.4% ±4.293 / 123
LiveBench Instruction FollowingInstruction62.8%50 / 66
LiveBench MathematicsMath84.3%53 / 66
LiveBench LanguageKnowledge73.6%54 / 66
SkillsBenchCoding40.6% ±4.629 / 35
Terminal-Bench 2.1 · Vals AIAgentic41.9% ±2.666 / 75
Code MigrationCoding6.8% ±1.268 / 77
LiveBench CodingCoding69.9%60 / 66
LiveBench ReasoningReasoning70.8%74 / 80
LiveBench Data AnalysisReasoning55.8%75 / 80
LiveBench Agentic CodingAgentic18.5%66 / 66
CaseLaw v2Legal79.3% ±1.41 / 53
CorpFin v2Finance68.5% ±0.98 / 131
LegalBenchLegal84.5% ±0.432 / 149
Finance Agent v1.1Finance53.8% ±2.818 / 51
ProgramBenchCoding0.0%34 / 65
TaxEval v2Finance70.8% ±0.986 / 143
MedCodeHealthcare38.1% ±2.172 / 106
MedScribeHealthcare74.4% ±2.082 / 109
Legal Research BenchLegal15.4% ±2.573 / 90
Finance Agent (v2)Finance37.7% ±0.466 / 79
Public Benefits Bench v1Public benefits50.1% ±1.311 / 13
MortgageTaxFinance48.3% ±1.085 / 97
Public Benefits Bench v1.1Public benefits51.7% ±1.345 / 50
Tax Agent BenchFinance6.2% ±1.776 / 82
EMBFinance18.4% ±1.770 / 75
ResearchClawBenchAgentic12.4%—

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseVals AINo licence statedLiveBenchNo licence statedmodels.devMIT

History

5 releases
Index54.3 → 53.3
23 Oct 202612 Nov 2026
Output / 1M$2.50 → $2.50
23 Oct 202612 Nov 2026

Questions

How much does Grok 4.3 cost?
$1.25 per 1M input tokens, $2.50 per 1M output tokens.
How good is Grok 4.3?
Its index is 53.3 out of 100. That is place 115 of 283 ranked models.
What is Grok 4.3 best at?
Instruction: 59.3 out of 100, place 15 of 63.
How fast is Grok 4.3?
61 output tokens per second, 1.3 s to the first token.
What is the context window of Grok 4.3?
1M tokens.
Is Grok 4.3 open source?
No. Its weights are not published.
When was Grok 4.3 released?
April 2026.

Badge

Grok 4.3 on Inference360: place 115 of 283, index 53.3
HTML
<a href="https://inference360.ai/models/grok-4-3"><img src="https://inference360.ai/badges/grok-4-3.svg" alt="Grok 4.3 on Inference360: place 115 of 283, index 53.3" height="20"></a>
Markdown
[![Grok 4.3 on Inference360: place 115 of 283, index 53.3](https://inference360.ai/badges/grok-4-3.svg)](https://inference360.ai/models/grok-4-3)

Compare with

GLM-5 Thinking53.4Muse Glimmer 30B53.0Grok Build 0.152.7dots3-note Preview63.4

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from xAI

All models
Grok Build 0.152.7Grok 4.2052.8Grok 4.20 Multi-agentUnrankedGrok 4.563.7Grok LatestUnranked