Enter opens the highlighted result. Esc closes.

Grok 4.3 vs Mistral Large 3 2512

Grok 4.3 leads Mistral Large 3 2512 on the Artificial Analysis Intelligence Index, 24.9 to 9.3, and Mistral Large 3 2512 costs 1.7x less per output token ($1.50 vs $2.50 per million). Updated October 21, 2026; 25 benchmarks have a result for both.

Replace Grok 4.3 Replace Mistral Large 3 2512
Grok 4.3xAI
Coding
37.4
Overall
24.9
Output $/M
$2.50
22
Mistral Large 3 2512Mistral AI
Coding
No data
Overall
9.3
Output $/M
$1.50
3
vs boards won of 25 shared

Grok 4.3 or Mistral Large 3 2512: which to pick

Only Grok 4.3 has a coding score, 37.4 on the Vibe Coding Index; Mistral Large 3 2512 has none, so the two cannot be ranked on coding.

Pick Mistral Large 3 2512 for output tokens 1.7x cheaper ($1.50 vs $2.50 per million) and work measured by TaxEval (Vals AI) or MortgageTax (Vals AI), where it leads.

Model AGrok 4.3xAI37.4/1003 independent benchmarks$0.31 per solved taskModel BMistral Large 3 2512Mistral AINot benchmarked1 independent benchmark$0.77 per solved task
Shared results
25
Grok 4.3 only
79
Mistral Large 3 2512 only
3
Like-for-like categories
7 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

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
Grok 4.3 37.4 from 3 benchmarks, Mistral Large 3 2512 not benchmarked.
Widest category
Coding, 26.0 points apart on average over 4 shared benchmarks.
Widest single gap
SWE-bench Verified (Vals AI): Grok 4.3 71.4% against Mistral Large 3 2512 41.4%, 30.0 points apart on a 0 to 100 scale.
Output price
Grok 4.3 $2.50 against Mistral Large 3 2512 $1.50 per million output tokens, 1.7x apart.
Context window
Grok 4.3 takes 1M tokens, Mistral Large 3 2512 262K.

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
CategoryGrok 4.3BarsMistral Large 3 2512Ties
Coding40
Intelligence10
Agentic work10
Reasoning20
Knowledge42
Web design80
General21
All shared boards223
CodingLike-for-likeSix of these boards feed the Vibe Coding Index40
Coding benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
LMArena WebDev1358Not measuredOnly Grok 4.3 measured
AA Coding Index42.220.1Grok 4.3
LiveCodeBench (Vals AI)84.5%55.3%Grok 4.3
SWE-bench Verified (Vals AI)71.4%41.4%Grok 4.3
Vibe Code Bench (Vals AI)19.4%Not measuredOnly Grok 4.3 measured
Code migration (Vals AI)6.8%Not measuredOnly Grok 4.3 measured
ProgramBench (Vals AI)0.0%Not measuredOnly Grok 4.3 measured
Terminal-Bench 2.1 (Vals AI)42.0%Not measuredOnly Grok 4.3 measured
SkillsBench (Vals AI)40.6%Not measuredOnly Grok 4.3 measured
SciCode48.3%Not measuredOnly Grok 4.3 measured
Terminal-Bench Hard (Artificial Analysis run)37.9%Not measuredOnly Grok 4.3 measured
WeirdML49.9%Not measuredOnly Grok 4.3 measured
ALE-Bench944.2264.7Grok 4.3
LMArena Text: coding1415Not measuredOnly Grok 4.3 measured
LMArena Text: software and IT1416Not measuredOnly Grok 4.3 measured
LMArena WebDev: React1353Not measuredOnly Grok 4.3 measured
LMArena WebDev: HTML1367Not measuredOnly Grok 4.3 measured
LMArena WebDev: from an image1375Not measuredOnly Grok 4.3 measured
Neither model has a result on the other 20 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
Intelligence benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
AA Intelligence Index24.99.3Grok 4.3
Neither model has a result on the other 1 board here
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
Agentic work benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
AA Agentic Index15.51.0Grok 4.3
Design Arena: agents building web apps1126Not measuredOnly Grok 4.3 measured
Design Arena: agents building full-stack apps1006Not measuredOnly Grok 4.3 measured
Design Arena: agents building mobile apps1059Not measuredOnly Grok 4.3 measured
Design Arena: agents building Android apps961Not measuredOnly Grok 4.3 measured
Design Arena: agents building Godot games914Not measuredOnly Grok 4.3 measured
Design Arena: agents building games1007Not measuredOnly Grok 4.3 measured
Design Arena: HTML slides1008Not measuredOnly Grok 4.3 measured
Design Arena: PowerPoint slides1013Not measuredOnly Grok 4.3 measured
Finance Agent (Vals AI)37.7%Not measuredOnly Grok 4.3 measured
Tax Agent Bench (Vals AI)6.2%Not measuredOnly Grok 4.3 measured
Legal Research Bench (Vals AI)15.4%Not measuredOnly Grok 4.3 measured
Excel Modeling Benchmark (Vals AI)18.4%Not measuredOnly Grok 4.3 measured
τ²-Bench Telecom97.7%Not measuredOnly Grok 4.3 measured
Vending-Bench 2$35Not measuredOnly Grok 4.3 measured
LMArena Search1166Not measuredOnly Grok 4.3 measured
Neither model has a result on the other 23 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index20
Reasoning benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
GPQA Diamond91.4%68.4%Grok 4.3
MMMU Pro (Vals AI)83.1%66.2%Grok 4.3
GPQA Diamond (Artificial Analysis run)90.1%Not measuredOnly Grok 4.3 measured
Humanity's Last Exam (Artificial Analysis run)37.2%Not measuredOnly Grok 4.3 measured
GPQA Diamond (Epoch AI run)88.8%Not measuredOnly Grok 4.3 measured
Chess puzzles18.0%Not measuredOnly Grok 4.3 measured
LMArena Text: hard prompts1396Not measuredOnly Grok 4.3 measured
LMArena Text: expert1384Not measuredOnly Grok 4.3 measured
LMArena Text: hard prompts in English1402Not measuredOnly Grok 4.3 measured
LMArena Vision: diagrams1241Not measuredOnly Grok 4.3 measured
LMArena Vision: homework1241Not measuredOnly Grok 4.3 measured
Neither model has a result on the other 10 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
Math benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
AIMENot measured42.9%Only Mistral Large 3 2512 measured
MGSM (Vals AI)Not measured85.4%Only Mistral Large 3 2512 measured
OTIS Mock AIME93.3%Not measuredOnly Grok 4.3 measured
FrontierMath, tiers 1 to 342.8%Not measuredOnly Grok 4.3 measured
FrontierMath, tier 414.6%Not measuredOnly Grok 4.3 measured
ProofBench11.0%Not measuredOnly Grok 4.3 measured
LMArena Text: math1387Not measuredOnly Grok 4.3 measured
LMArena Text: mathematical work1396Not measuredOnly Grok 4.3 measured
Neither model has a result on the other 5 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index42
Knowledge benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
MMLU Pro85.8%79.8%Grok 4.3
LegalBench (Vals AI)84.5%79.1%Grok 4.3
CaseLaw (Vals AI)79.3%61.4%Grok 4.3
CorpFin (Vals AI)68.5%61.0%Grok 4.3
TaxEval (Vals AI)70.8%73.1%Mistral Large 3 2512
MortgageTax (Vals AI)48.3%52.1%Mistral Large 3 2512
MedQA (Vals AI)Not measured82.2%Only Mistral Large 3 2512 measured
MedCode (Vals AI)38.1%Not measuredOnly Grok 4.3 measured
SimpleQA Verified33.2%Not measuredOnly Grok 4.3 measured
LMArena Text: science1399Not measuredOnly Grok 4.3 measured
LMArena Text: medicine and healthcare1384Not measuredOnly Grok 4.3 measured
LMArena Text: legal and government1386Not measuredOnly Grok 4.3 measured
LMArena Text: business and finance1381Not measuredOnly Grok 4.3 measured
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index80
Web design benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
Design Arena: websites11961162Grok 4.3
Design Arena: UI components11921102Grok 4.3
Design Arena: game dev11741080Grok 4.3
Design Arena: data viz11791136Grok 4.3
Design Arena: 3D11251100Grok 4.3
Design Arena: all code categories11871142Grok 4.3
Design Arena: SVG1081993Grok 4.3
Design Arena: ASCII art11251069Grok 4.3
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index21
General benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
Public Benefits Bench (Vals AI)51.7%Not measuredOnly Grok 4.3 measured
SAGE (Vals AI)19.7%24.6%Mistral Large 3 2512
MedScribe (Vals AI)74.4%Not measuredOnly Grok 4.3 measured
Vals Multimodal Index43.3%Not measuredOnly Grok 4.3 measured
IFBench83.3%Not measuredOnly Grok 4.3 measured
Long context reasoning (AA-LCR)75.0%Not measuredOnly Grok 4.3 measured
DTBench90.7%65.1%Grok 4.3
LMCA38.3%16.7%Grok 4.3
LMArena Text1397Not measuredOnly Grok 4.3 measured
LMArena Text: instruction following1366Not measuredOnly Grok 4.3 measured
LMArena Text: creative writing1380Not measuredOnly Grok 4.3 measured
LMArena Vision1229Not measuredOnly Grok 4.3 measured
LMArena Text: English1404Not measuredOnly Grok 4.3 measured
LMArena Text: non-English1385Not measuredOnly Grok 4.3 measured
LMArena Text: Chinese1422Not measuredOnly Grok 4.3 measured
LMArena Text: German1395Not measuredOnly Grok 4.3 measured
LMArena Text: Japanese1379Not measuredOnly Grok 4.3 measured
LMArena Text: Korean1354Not measuredOnly Grok 4.3 measured
LMArena Text: French1411Not measuredOnly Grok 4.3 measured
LMArena Text: Spanish1399Not measuredOnly Grok 4.3 measured
LMArena Text: Russian1399Not measuredOnly Grok 4.3 measured
LMArena Text: Polish1398Not measuredOnly Grok 4.3 measured
LMArena Text: multi-turn1406Not measuredOnly Grok 4.3 measured
LMArena Text: longer queries1393Not measuredOnly Grok 4.3 measured
LMArena Text: writing and language1385Not measuredOnly Grok 4.3 measured
LMArena Text: entertainment and media1370Not measuredOnly Grok 4.3 measured
LMArena Vision: English1235Not measuredOnly Grok 4.3 measured
LMArena Vision: Chinese1257Not measuredOnly Grok 4.3 measured
LMArena Vision: reading text1235Not measuredOnly Grok 4.3 measured
LMArena Vision: humor1222Not measuredOnly Grok 4.3 measured
LMArena Vision: recognizing things1217Not measuredOnly Grok 4.3 measured
LMArena Vision: creative writing1237Not measuredOnly Grok 4.3 measured
Neither model has a result on the other 14 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Grok 4.3 and Mistral Large 3 2512
BenchmarkGrok 4.3Mistral Large 3 2512Leader
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

Grok 4.3$0.0025Fits in one request
Mistral Large 3 2512$0.0013Fits in one request

Mistral Large 3 2512 costs less

Repository review

50K input + 3K output tokens

Grok 4.3$0.070Fits in one request
Mistral Large 3 2512$0.029Fits in one request

Mistral Large 3 2512 costs less

Agent loop

200K input + 10K output tokens

Grok 4.3$0.275Fits in one request
Mistral Large 3 2512$0.115Fits in one request

Mistral Large 3 2512 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 Grok 4.3 and Mistral Large 3 2512
SpecificationSpaceXAI Grok 4.3Mistral Large 3 2512
MakerxAIMistral AI
WeightsNo public weights listedNo public weights listed
ReleasedApril 30, 2026December 1, 2025
Context window1M262K
Input $/Mtok$1.25$0.50
Output $/Mtok$2.50$1.50
Coding index37.4Not benchmarked
Independent measurements behind it31
Cost per solved task$0.31$0.77
SourceOpenRouter listingOpenRouter listing

Questions

Is Grok 4.3 better than Mistral Large 3 2512 for coding?

Only Grok 4.3 has a coding score, 37.4 on the Vibe Coding Index; Mistral Large 3 2512 has none, so the two cannot be ranked on coding. Pick Mistral Large 3 2512 for output tokens 1.7x cheaper ($1.50 vs $2.50 per million) and work measured by TaxEval (Vals AI) or MortgageTax (Vals AI), where it leads.

Which is cheaper, Grok 4.3 or Mistral Large 3 2512?

Grok 4.3 costs $1.25 per million input tokens and $2.50 per million output tokens; Mistral Large 3 2512 costs $0.50 and $1.50. Per solved SWE-bench Verified task, Grok 4.3 costs $0.31 and Mistral Large 3 2512 $0.77.

Which has the bigger context window, Grok 4.3 or Mistral Large 3 2512?

Grok 4.3, with 1M tokens against 262K.

How many benchmarks is the Grok 4.3 vs Mistral Large 3 2512 comparison based on?

25 benchmarks have a result for both models as of October 21, 2026. Grok 4.3's coding index rests on 3 independent benchmarks and Mistral Large 3 2512's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next