Enter opens the highlighted result. Esc closes.

DeepSeek V3.2 Exp vs Llama 4 Maverick

DeepSeek V3.2 Exp leads Llama 4 Maverick on the Artificial Analysis Intelligence Index, 16.6 to 9.3, and DeepSeek V3.2 Exp costs 2.4x less per output token ($0.41 vs $1.00 per million). Updated October 21, 2026; 52 benchmarks have a result for both.

Replace DeepSeek V3.2 Exp Replace Llama 4 Maverick
DeepSeek V3.2 ExpDeepSeek
Coding
No data
Overall
No data
Output $/M
$0.41
51
Llama 4 MaverickMeta Llama
Coding
No data
Overall
21.8
Output $/M
$1.00
1
vs boards won of 52 shared

DeepSeek V3.2 Exp or Llama 4 Maverick: which to pick

Neither model has a coding score yet, so the two cannot be ranked on coding.

Pick DeepSeek V3.2 Exp for output tokens 2.4x cheaper ($0.41 vs $1.00 per million) and work measured by Kagi LLM Benchmark, where it leads.

Model ADeepSeek V3.2 ExpDeepSeekNot benchmarked1 independent benchmarkNo cost per solved taskModel BLlama 4 MaverickMeta LlamaNot benchmarkedNo coding benchmark$1.491 per solved task
Shared results
52
DeepSeek V3.2 Exp only
7
Llama 4 Maverick only
33
Like-for-like categories
8 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
DeepSeek V3.2 Exp not benchmarked, Llama 4 Maverick not benchmarked.
Widest category
Math, 45.1 points apart on average over 4 shared benchmarks.
Widest single gap
AA Math Index: DeepSeek V3.2 Exp 87.7 against Llama 4 Maverick 19.3, 68.4 points apart on a 0 to 100 scale.
Output price
DeepSeek V3.2 Exp $0.41 against Llama 4 Maverick $1.00 per million output tokens, 2.4x apart.
Context window
DeepSeek V3.2 Exp takes 164K tokens, Llama 4 Maverick 524K.

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
CategoryDeepSeek V3.2 ExpBarsLlama 4 MaverickTies
Coding60
Intelligence20
Agentic work10
Reasoning51
Math40
Knowledge50
Web design60
General220
All shared boards511
CodingLike-for-likeSix of these boards feed the Vibe Coding Index60
Coding benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
Aider polyglot74.2%15.6%DeepSeek V3.2 Exp
SWE-bench VerifiedNot measured21.0%Only Llama 4 Maverick measured
LMArena WebDev1272Not measuredOnly DeepSeek V3.2 Exp measured
AA Coding IndexNot measured16.3Only Llama 4 Maverick measured
LiveCodeBench (Vals AI)Not measured47.3%Only Llama 4 Maverick measured
LiveCodeBench (Artificial Analysis run)78.9%39.7%DeepSeek V3.2 Exp
SciCodeNot measured31.7%Only Llama 4 Maverick measured
Terminal-Bench Hard (Artificial Analysis run)31.1%6.8%DeepSeek V3.2 Exp
WeirdML39.5%24.5%DeepSeek V3.2 Exp
ALE-BenchNot measured173.0Only Llama 4 Maverick measured
LMArena Text: coding14381301DeepSeek V3.2 Exp
LMArena Text: software and IT14381299DeepSeek V3.2 Exp
LMArena WebDev: HTML1281Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 26 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
AA Intelligence Index16.69.3DeepSeek V3.2 Exp
Epoch Capabilities Index144.9132.2DeepSeek V3.2 Exp
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
Agentic work benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
AA Agentic IndexNot measured0.6Only Llama 4 Maverick measured
τ²-Bench Telecom33.9%17.8%DeepSeek V3.2 Exp
TheAgentCompany42.9%Not measuredOnly DeepSeek V3.2 Exp measured
Berkeley Function Calling Leaderboard (BFCL)56.7%Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 37 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index51
Reasoning benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
GPQA DiamondNot measured69.4%Only Llama 4 Maverick measured
MMMU Pro (Vals AI)Not measured71.7%Only Llama 4 Maverick measured
GPQA Diamond (Artificial Analysis run)79.7%67.1%DeepSeek V3.2 Exp
Humanity's Last Exam (Artificial Analysis run)14.9%4.9%DeepSeek V3.2 Exp
ARC-AGINot measured4.4%Only Llama 4 Maverick measured
ARC-AGI-2Not measured0.0%Only Llama 4 Maverick measured
SimpleBenchNot measured27.7%Only Llama 4 Maverick measured
LMArena Text: hard prompts14291281DeepSeek V3.2 Exp
LMArena Text: expert14221259DeepSeek V3.2 Exp
LMArena Text: hard prompts in English14441295DeepSeek V3.2 Exp
LMArena Vision: diagramsNot measured1134Only Llama 4 Maverick measured
LMArena Vision: homeworkNot measured1164Only Llama 4 Maverick measured
Kagi LLM Benchmark52.2%55.9%Llama 4 Maverick
Neither model has a result on the other 9 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index40
Math benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
AIMENot measured25.2%Only Llama 4 Maverick measured
MATH 500 (Vals AI)Not measured85.2%Only Llama 4 Maverick measured
MGSM (Vals AI)Not measured92.4%Only Llama 4 Maverick measured
AA Math Index87.719.3DeepSeek V3.2 Exp
AIME 2025 (Artificial Analysis run)87.7%19.3%DeepSeek V3.2 Exp
LMArena Text: math14221299DeepSeek V3.2 Exp
LMArena Text: mathematical work14331295DeepSeek V3.2 Exp
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index50
Knowledge benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
MMLU ProNot measured79.4%Only Llama 4 Maverick measured
LegalBench (Vals AI)Not measured77.8%Only Llama 4 Maverick measured
CorpFin (Vals AI)Not measured49.7%Only Llama 4 Maverick measured
TaxEval (Vals AI)Not measured66.6%Only Llama 4 Maverick measured
MortgageTax (Vals AI)Not measured58.5%Only Llama 4 Maverick measured
MedQA (Vals AI)Not measured43.3%Only Llama 4 Maverick measured
MedCode (Vals AI)Not measured36.5%Only Llama 4 Maverick measured
MMLU Pro (Artificial Analysis run)85.0%80.9%DeepSeek V3.2 Exp
GeoBenchNot measured52.0%Only Llama 4 Maverick measured
LMArena Text: science14431288DeepSeek V3.2 Exp
LMArena Text: medicine and healthcare14681275DeepSeek V3.2 Exp
LMArena Text: legal and government14191286DeepSeek V3.2 Exp
LMArena Text: business and finance14121260DeepSeek V3.2 Exp
Neither model has a result on the other 5 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index60
Web design benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
Design Arena: websites1179871DeepSeek V3.2 Exp
Design Arena: UI components1167899DeepSeek V3.2 Exp
Design Arena: game dev1144838DeepSeek V3.2 Exp
Design Arena: data viz1152882DeepSeek V3.2 Exp
Design Arena: 3D1148901DeepSeek V3.2 Exp
Design Arena: all code categories1171880DeepSeek V3.2 Exp
Design Arena: SVG1032Not measuredOnly DeepSeek V3.2 Exp measured
Neither model has a result on the other 2 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index220
General benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
SAGE (Vals AI)Not measured35.7%Only Llama 4 Maverick measured
MedScribe (Vals AI)Not measured54.2%Only Llama 4 Maverick measured
IFBench54.1%43.0%DeepSeek V3.2 Exp
Long context reasoning (AA-LCR)72.3%50.0%DeepSeek V3.2 Exp
DTBench87.7%61.9%DeepSeek V3.2 Exp
LMCA29.1%15.9%DeepSeek V3.2 Exp
Fiction.LiveBench83.3%46.2%DeepSeek V3.2 Exp
Creative writing (Lech Mazur)Not measured62.0%Only Llama 4 Maverick measured
CL-bench13.2%Not measuredOnly DeepSeek V3.2 Exp measured
CL-bench Life9.5%Not measuredOnly DeepSeek V3.2 Exp measured
LMArena Text14251287DeepSeek V3.2 Exp
LMArena Text: instruction following14011267DeepSeek V3.2 Exp
LMArena Text: creative writing14041266DeepSeek V3.2 Exp
LMArena VisionNot measured1142Only Llama 4 Maverick measured
LMArena Text: English14411303DeepSeek V3.2 Exp
LMArena Text: non-English14041269DeepSeek V3.2 Exp
LMArena Text: Chinese14611277DeepSeek V3.2 Exp
LMArena Text: German14401291DeepSeek V3.2 Exp
LMArena Text: Japanese13311206DeepSeek V3.2 Exp
LMArena Text: Korean13721203DeepSeek V3.2 Exp
LMArena Text: French14341259DeepSeek V3.2 Exp
LMArena Text: Spanish14411293DeepSeek V3.2 Exp
LMArena Text: Russian14121286DeepSeek V3.2 Exp
LMArena Text: Polish14351275DeepSeek V3.2 Exp
LMArena Text: multi-turn14221289DeepSeek V3.2 Exp
LMArena Text: longer queries14181280DeepSeek V3.2 Exp
LMArena Text: writing and language14071277DeepSeek V3.2 Exp
LMArena Text: entertainment and media14071250DeepSeek V3.2 Exp
LMArena Vision: EnglishNot measured1150Only Llama 4 Maverick measured
LMArena Vision: ChineseNot measured1141Only Llama 4 Maverick measured
LMArena Vision: reading textNot measured1143Only Llama 4 Maverick measured
LMArena Vision: humorNot measured1100Only Llama 4 Maverick measured
LMArena Vision: creative writingNot measured1099Only Llama 4 Maverick measured
Neither model has a result on the other 13 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V3.2 Exp and Llama 4 Maverick
BenchmarkDeepSeek V3.2 ExpLlama 4 MaverickLeader
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

DeepSeek V3.2 Exp$0.0005Fits in one request
Llama 4 Maverick$0.0008Fits in one request

DeepSeek V3.2 Exp costs less

Repository review

50K input + 3K output tokens

DeepSeek V3.2 Exp$0.015Fits in one request
Llama 4 Maverick$0.021Fits in one request

DeepSeek V3.2 Exp costs less

Agent loop

200K input + 10K output tokens

DeepSeek V3.2 Exp$0.058Exceeds the context window
Llama 4 Maverick$0.080Fits in one request

DeepSeek V3.2 Exp 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 DeepSeek V3.2 Exp and Llama 4 Maverick
SpecificationDeepSeek V3.2 ExpMeta Llama 4 Maverick
MakerDeepSeekMeta Llama
WeightsOpen weightsOpen weights
ReleasedSeptember 29, 2025April 5, 2025
Context window164K524K
Input $/Mtok$0.27$0.35
Output $/Mtok$0.41$1.00
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it10
Cost per solved taskNot computable$1.491
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V3.2 Exp better than Llama 4 Maverick for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick DeepSeek V3.2 Exp for output tokens 2.4x cheaper ($0.41 vs $1.00 per million) and work measured by Kagi LLM Benchmark, where it leads.

Which is cheaper, DeepSeek V3.2 Exp or Llama 4 Maverick?

DeepSeek V3.2 Exp costs $0.27 per million input tokens and $0.41 per million output tokens; Llama 4 Maverick costs $0.35 and $1.00. Cost per solved task is known for Llama 4 Maverick only.

Which has the bigger context window, DeepSeek V3.2 Exp or Llama 4 Maverick?

Llama 4 Maverick, with 524K tokens against 164K.

How many benchmarks is the DeepSeek V3.2 Exp vs Llama 4 Maverick comparison based on?

52 benchmarks have a result for both models as of October 21, 2026. DeepSeek V3.2 Exp's coding index rests on 1 independent benchmark and Llama 4 Maverick's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next