LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

GPT-5.4 vs Grok 4.6

Updated October 21, 2026. 93 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace GPT-5.4 Replace Grok 4.6
GPT-5.4OpenAI
Coding
37.7
Overall
39.0
Output $/M
$15.00
52
Grok 4.6xAI
Coding
76.4
Overall
44.3
Output $/M
$6.00
41
vs boards won of 93 shared

Decision reading

Pick Grok 4.6 for coding: it scores 76.4 on the Vibe Coding Index against 37.7 for GPT-5.4.

Pick GPT-5.4 when the context window is larger, 1.05M against 500K tokens, or when what matters is WeirdML or ALE-Bench, where it leads.

Model AGPT-5.4OpenAI37.7/1001 independent benchmark$1.02 per solved taskModel BGrok 4.6xAI76.4/1002 independent benchmarks$0.82 per solved task
Shared results
93
GPT-5.4 only
16
Grok 4.6 only
30
Like-for-like categories
8 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
GPT-5.4 37.7 from 1 benchmark, Grok 4.6 76.4 from 2 benchmarks.
Widest category
Agentic work, 22.0 points apart on average over 7 shared benchmarks.
Widest single gap
Design Arena: agents building Android apps: GPT-5.4 1037 against Grok 4.6 1297, 32.5 points apart on a 0 to 100 scale.
Output price
GPT-5.4 $15.00 against Grok 4.6 $6.00 per million output tokens, 2.5x apart.
Context window
GPT-5.4 takes 1.05M tokens, Grok 4.6 500K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryGPT-5.4BarsGrok 4.6Ties
Coding411
Intelligence11
Agentic work07
Reasoning95
Math51
Knowledge65
Web design08
General273
All shared boards5241
CodingLike-for-likeFour of these boards feed the Vibe Coding Index411
Coding benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
LiveBench codingNot measured89.1%Only Grok 4.6 measured
LMArena WebDev14641617Grok 4.6
AA Coding Index71.176.8Grok 4.6
LiveCodeBench (Vals AI)84.1%88.2%Grok 4.6
SWE-bench Verified (Vals AI)78.2%95.6%Grok 4.6
Terminal-Bench 4 (Vals AI)Not measured17.2%Only Grok 4.6 measured
Vibe Code Bench (Vals AI)67.4%76.2%Grok 4.6
Code migration (Vals AI)35.0%44.6%Grok 4.6
ProgramBench (Vals AI)0.5%1.0%Grok 4.6
IOI (Vals AI)Not measured47.6%Only Grok 4.6 measured
Terminal-Bench 2.1 (Vals AI)Not measured78.3%Only Grok 4.6 measured
Vibe Code Bench 1-100 (Vals AI)Not measured14.8%Only Grok 4.6 measured
SkillsBench (Vals AI)51.7%55.8%Grok 4.6
CyberBench (Vals AI)Not measured67.7%Only Grok 4.6 measured
SciCodeNot measured56.5%Only Grok 4.6 measured
Terminal-Bench Hard (Artificial Analysis run)57.6%Not measuredOnly GPT-5.4 measured
Terminal-Bench81.8%Not measuredOnly GPT-5.4 measured
FrontierCodeNot measured48.0%Only Grok 4.6 measured
FrontierSWENot measured25.3%Only Grok 4.6 measured
GSO-Bench31.4%Not measuredOnly GPT-5.4 measured
WeirdML77.7%67.3%GPT-5.4
MirrorCode15.6%Not measuredOnly GPT-5.4 measured
ALE-Bench1607.01508.1GPT-5.4
LMArena Text: coding14961466GPT-5.4
LMArena Text: software and IT14911457GPT-5.4
LMArena WebDev: React14541631Grok 4.6
LMArena WebDev: HTML14781597Grok 4.6
LMArena WebDev: from an image14361603Grok 4.6
Neither model has a result on the other 10 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
Intelligence benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
AA Intelligence Index39.044.3Grok 4.6
Epoch Capabilities Index157.3156.6GPT-5.4
Agentic workLike-for-likePublic boards, shown as published, kept out of the index07
Agentic work benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
AA Agentic IndexNot measured53.0Only Grok 4.6 measured
Design Arena: agents building web apps10591247Grok 4.6
Design Arena: agents building full-stack apps10171264Grok 4.6
Design Arena: agents building mobile apps10871234Grok 4.6
Design Arena: agents building Android apps10371297Grok 4.6
Design Arena: agents building Godot games11351276Grok 4.6
Design Arena: agents building gamesNot measured1208Only Grok 4.6 measured
Design Arena: HTML slidesNot measured1249Only Grok 4.6 measured
Design Arena: PowerPoint slidesNot measured1241Only Grok 4.6 measured
LiveBench researchNot measured84.5%Only Grok 4.6 measured
LiveBench browserNot measured64.0%Only Grok 4.6 measured
LiveBench long agenticNot measured40.7%Only Grok 4.6 measured
Finance Agent (Vals AI)Not measured53.7%Only Grok 4.6 measured
Tax Agent Bench (Vals AI)Not measured33.1%Only Grok 4.6 measured
Legal Research Bench (Vals AI)Not measured48.1%Only Grok 4.6 measured
Excel Modeling Benchmark (Vals AI)Not measured62.7%Only Grok 4.6 measured
BioMysteryBench (Vals AI)Not measured72.2%Only Grok 4.6 measured
Terminal-Bench-Science (Vals AI)Not measured11.4%Only Grok 4.6 measured
Vals Science IndexNot measured38.1%Only Grok 4.6 measured
Time Horizon Index (Vals AI)Not measured5.8%Only Grok 4.6 measured
τ²-Bench Telecom87.1%Not measuredOnly GPT-5.4 measured
APEX-Agents52.4%65.3%Grok 4.6
METR time horizons74.3%Not measuredOnly GPT-5.4 measured
DeepResearch Bench35.1%Not measuredOnly GPT-5.4 measured
PostTrainBench19.0%Not measuredOnly GPT-5.4 measured
Vending-Bench 2$6,144$9,047Grok 4.6
Neither model has a result on the other 13 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index95
Reasoning benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
GPQA Diamond91.7%94.7%Grok 4.6
LiveBench reasoningNot measured91.2%Only Grok 4.6 measured
MysteryMechanism (Vals AI)Not measured30.6%Only Grok 4.6 measured
MMMU Pro (Vals AI)87.5%Not measuredOnly GPT-5.4 measured
GPQA Diamond (Artificial Analysis run)92.0%94.9%Grok 4.6
Humanity's Last Exam (Artificial Analysis run)43.7%44.1%Grok 4.6
GPQA Diamond (Epoch AI run)93.3%94.0%Grok 4.6
ARC-AGI93.7%87.5%GPT-5.4
ARC-AGI-274.0%67.1%GPT-5.4
SimpleBenchNot measured75.9%Only Grok 4.6 measured
Chess puzzles48.0%33.0%GPT-5.4
Mystery game puzzles37.0%34.0%GPT-5.4
EBR-bench25.4%30.5%Grok 4.6
LMArena Text: hard prompts14851448GPT-5.4
LMArena Text: expert15071469GPT-5.4
LMArena Text: hard prompts in English14791445GPT-5.4
LMArena Vision: diagrams13271285GPT-5.4
LMArena Vision: homework13361283GPT-5.4
Neither model has a result on the other 3 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index51
Math benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
AIME96.7%Not measuredOnly GPT-5.4 measured
OTIS Mock AIME97.8%99.2%Grok 4.6
FrontierMath, tiers 1 to 378.6%66.0%GPT-5.4
FrontierMath, tier 449.0%31.7%GPT-5.4
ProofBench56.0%51.0%GPT-5.4
LMArena Text: math14881417GPT-5.4
LMArena Text: mathematical work14971464GPT-5.4
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index65
Knowledge benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
MMLU Pro87.5%89.4%Grok 4.6
LegalBench (Vals AI)86.0%86.3%Grok 4.6
CaseLaw (Vals AI)63.8%Not measuredOnly GPT-5.4 measured
CorpFin (Vals AI)65.3%66.2%Grok 4.6
TaxEval (Vals AI)74.0%71.1%GPT-5.4
MortgageTax (Vals AI)68.3%64.2%GPT-5.4
MedQA (Vals AI)96.1%Not measuredOnly GPT-5.4 measured
MedCode (Vals AI)41.3%44.7%Grok 4.6
SimpleQA Verified45.1%49.3%Grok 4.6
LMArena Text: science14791437GPT-5.4
LMArena Text: medicine and healthcare14581420GPT-5.4
LMArena Text: legal and government14861429GPT-5.4
LMArena Text: business and finance14741416GPT-5.4
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index08
Web design benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
Design Arena: websites11991289Grok 4.6
Design Arena: UI components12151294Grok 4.6
Design Arena: game dev11951296Grok 4.6
Design Arena: data viz11961283Grok 4.6
Design Arena: 3D10651277Grok 4.6
Design Arena: all code categories11801292Grok 4.6
Design Arena: SVG11411239Grok 4.6
Design Arena: ASCII art11881278Grok 4.6
LiveBench designNot measured79.8%Only Grok 4.6 measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index273
General benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
LiveBench data analysisNot measured75.3%Only Grok 4.6 measured
Public Benefits Bench (Vals AI)Not measured66.8%Only Grok 4.6 measured
SAGE (Vals AI)43.3%28.9%GPT-5.4
MedScribe (Vals AI)77.5%86.5%Grok 4.6
IFBench73.9%Not measuredOnly GPT-5.4 measured
Long context reasoning (AA-LCR)82.0%81.0%GPT-5.4
DTBench94.4%97.3%Grok 4.6
LMCA52.0%48.5%GPT-5.4
CL-bench27.9%Not measuredOnly GPT-5.4 measured
Furniture assembly37.5%40.0%Grok 4.6
CL-bench Life21.7%Not measuredOnly GPT-5.4 measured
LMArena Text14691429GPT-5.4
LMArena Text: instruction following14691431GPT-5.4
LMArena Text: creative writing14391426GPT-5.4
LMArena Vision13031263GPT-5.4
LMArena Text: English14671433GPT-5.4
LMArena Text: non-English14651420GPT-5.4
LMArena Text: Chinese15201476GPT-5.4
LMArena Text: German14721426GPT-5.4
LMArena Text: Japanese14851378GPT-5.4
LMArena Text: Korean14491391GPT-5.4
LMArena Text: French14931455GPT-5.4
LMArena Text: Spanish14521414GPT-5.4
LMArena Text: Russian14801421GPT-5.4
LMArena Text: Polish14721430GPT-5.4
LMArena Text: multi-turn14821425GPT-5.4
LMArena Text: longer queries14731456GPT-5.4
LMArena Text: writing and language14581429GPT-5.4
LMArena Text: entertainment and media14381421GPT-5.4
LMArena Vision: English13011270GPT-5.4
LMArena Vision: Chinese13541289GPT-5.4
LMArena Vision: reading text13121272GPT-5.4
LMArena Vision: humor12841270GPT-5.4
LMArena Vision: recognizing things1251Not measuredOnly GPT-5.4 measured
LMArena Vision: creative writing12831262GPT-5.4
LMArena Document14711452GPT-5.4
Neither model has a result on the other 10 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5.4 and Grok 4.6
BenchmarkGPT-5.4Grok 4.6Leader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

GPT-5.4$0.010Fits in one request
Grok 4.6$0.0050Fits in one request

Grok 4.6 costs less

Repository review

50K input + 3K output tokens

GPT-5.4$0.170Fits in one request
Grok 4.6$0.118Fits in one request

Grok 4.6 costs less

Agent loop

200K input + 10K output tokens

GPT-5.4$0.650Fits in one request
Grok 4.6$0.460Fits in one request

Grok 4.6 costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of GPT-5.4 and Grok 4.6
SpecificationOpenAI GPT-5.4SpaceXAI Grok 4.6
MakerOpenAIxAI
WeightsNo public weights listedNo public weights listed
ReleasedMarch 5, 2026August 12, 2026
Context window1.05M500K
Input $/Mtok$2.50$2.00
Output $/Mtok$15.00$6.00
Coding index37.776.4
Independent measurements behind it12
Cost per solved task$1.02$0.82
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5.4 better than Grok 4.6 for coding?

Pick Grok 4.6 for coding: it scores 76.4 on the Vibe Coding Index against 37.7 for GPT-5.4. Pick GPT-5.4 when the context window is larger, 1.05M against 500K tokens, or when what matters is WeirdML or ALE-Bench, where it leads.

Which is cheaper, GPT-5.4 or Grok 4.6?

GPT-5.4 costs $2.50 per million input tokens and $15.00 per million output tokens; Grok 4.6 costs $2.00 and $6.00. Per solved SWE-bench Verified task, GPT-5.4 costs $1.02 and Grok 4.6 $0.82.

Which has the bigger context window, GPT-5.4 or Grok 4.6?

GPT-5.4, with 1.05M tokens against 500K.

How many benchmarks is the GPT-5.4 vs Grok 4.6 comparison based on?

93 benchmarks have a result for both models as of October 21, 2026. GPT-5.4's coding index rests on 1 independent benchmark and Grok 4.6's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next