Enter opens the highlighted result. Esc closes.

Qwen3.8 Max (0902) vs Qwen4 Max

Qwen4 Max leads Qwen3.8 Max (0902) on coding, 74.0 to 67.2 on the Vibe Coding Index, and Qwen4 Max costs 1.4x less per output token ($4.275 vs $6.00 per million). Updated November 22, 2026; 26 benchmarks have a result for both.

Qwen3.8 Max (0902)Qwen
Coding
67.2
Overall
66.3
Output $/M
$6.00
2
Qwen4 MaxQwen
Coding
74.0
Overall
66.6
Output $/M
$4.275
24
vs boards won of 26 shared

Qwen3.8 Max (0902) or Qwen4 Max: which to pick

Pick Qwen4 Max for coding: it scores 74.0 on the Vibe Coding Index against 67.2 for Qwen3.8 Max (0902), though on 4 benchmarks against 5.

Pick Qwen3.8 Max (0902) for work measured by Tax Agent Bench (Vals AI) or Legal Research Bench (Vals AI), where it leads.

Shared results
26
Qwen3.8 Max (0902) only
101
Qwen4 Max only
0
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
Qwen3.8 Max (0902) 67.2 from 5 benchmarks, Qwen4 Max 74.0 from 4 benchmarks.
Widest category
Agentic work, 14.1 points apart on average over 6 shared benchmarks.
Widest single gap
LiveBench browser: Qwen3.8 Max (0902) 49.4% against Qwen4 Max 80.7%, 31.3 points apart on a 0 to 100 scale.
Output price
Qwen3.8 Max (0902) $6.00 against Qwen4 Max $4.275 per million output tokens, 1.4x apart.
Context window
Both take 1M tokens.

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
CategoryQwen3.8 Max (0902)BarsQwen4 MaxTies
Coding011
Intelligence01
Agentic work24
Reasoning03
Knowledge01
Web design01
General03
All shared boards224
CodingLike-for-likeSix of these boards feed the Vibe Coding Index011
Coding benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
LiveBench coding60.9%69.8%Qwen4 Max
LMArena WebDev16741751Qwen4 Max
LiveCodeBench (Vals AI)87.8%Not measuredOnly Qwen3.8 Max (0902) measured
SWE-bench Verified (Vals AI)85.6%Not measuredOnly Qwen3.8 Max (0902) measured
Terminal-Bench 4 (Vals AI)38.9%48.5%Qwen4 Max
Vibe Code Bench (Vals AI)77.1%86.7%Qwen4 Max
Code migration (Vals AI)24.0%Not measuredOnly Qwen3.8 Max (0902) measured
ProgramBench (Vals AI)0.0%Not measuredOnly Qwen3.8 Max (0902) measured
IOI (Vals AI)68.9%Not measuredOnly Qwen3.8 Max (0902) measured
Terminal-Bench 2.1 (Vals AI)67.4%Not measuredOnly Qwen3.8 Max (0902) measured
SWE-bench Multimodal (Vals AI)42.5%Not measuredOnly Qwen3.8 Max (0902) measured
Vibe Code Bench 1-100 (Vals AI)24.8%Not measuredOnly Qwen3.8 Max (0902) measured
SkillsBench (Vals AI)42.0%Not measuredOnly Qwen3.8 Max (0902) measured
CyberBench (Vals AI)28.6%Not measuredOnly Qwen3.8 Max (0902) measured
SciCode53.2%Not measuredOnly Qwen3.8 Max (0902) measured
Terminal-Bench 4.0 (Artificial Analysis run)38.9%48.5%Qwen4 Max
FrontierCode 2.023.4%26.5%Qwen4 Max
FrontierSWE17.8%Not measuredOnly Qwen3.8 Max (0902) measured
MirrorCode52.2%Not measuredOnly Qwen3.8 Max (0902) measured
ALE-Bench1679.11831.0Qwen4 Max
LMArena Text: coding1503Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: software and IT1502Not measuredOnly Qwen3.8 Max (0902) measured
LMArena WebDev: React16961785Qwen4 Max
LMArena WebDev: HTML16531752Qwen4 Max
LMArena WebDev: from an image16391675Qwen4 Max
GameBench 269.782.4Qwen4 Max
Terminal-Bench 4.027.0%Not measuredOnly Qwen3.8 Max (0902) measured
Neither model has a result on the other 25 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
AA Intelligence Index33.742.9Qwen4 Max
Epoch Capabilities Index157.6Not measuredOnly Qwen3.8 Max (0902) measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index24
Agentic work benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
Design Arena: agents building web apps1270Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: agents building full-stack apps1282Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: agents building mobile apps1224Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: agents building Android apps1191Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: agents building Godot games1210Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: PowerPoint slides1243Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: spreadsheets1004Not measuredOnly Qwen3.8 Max (0902) measured
LiveBench research72.8%84.7%Qwen4 Max
LiveBench browser49.4%80.7%Qwen4 Max
LiveBench long agentic26.6%46.6%Qwen4 Max
Finance Agent (Vals AI)57.9%Not measuredOnly Qwen3.8 Max (0902) measured
Tax Agent Bench (Vals AI)49.7%40.4%Qwen3.8 Max (0902)
Legal Research Bench (Vals AI)47.6%41.4%Qwen3.8 Max (0902)
Harvey's Legal Agent Benchmark (Vals AI)15.4%21.3%Qwen4 Max
Excel Modeling Benchmark (Vals AI)72.0%Not measuredOnly Qwen3.8 Max (0902) measured
Terminal-Bench-Science (Vals AI)12.9%Not measuredOnly Qwen3.8 Max (0902) measured
Robo-bench (Vals AI)5.2%Not measuredOnly Qwen3.8 Max (0902) measured
APEX-Agents63.3%Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Search1261Not measuredOnly Qwen3.8 Max (0902) measured
τ³-Banking55.2%Not measuredOnly Qwen3.8 Max (0902) measured
Neither model has a result on the other 24 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index03
Reasoning benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
GPQA Diamond93.7%Not measuredOnly Qwen3.8 Max (0902) measured
LiveBench reasoning88.2%92.2%Qwen4 Max
MysteryMechanism (Vals AI)23.9%Not measuredOnly Qwen3.8 Max (0902) measured
MMMU Pro (Vals AI)88.0%Not measuredOnly Qwen3.8 Max (0902) measured
GPQA Diamond (Artificial Analysis run)92.8%Not measuredOnly Qwen3.8 Max (0902) measured
Humanity's Last Exam (Artificial Analysis run)43.1%Not measuredOnly Qwen3.8 Max (0902) measured
HLE-Diamond (Artificial Analysis run)24.0%44.9%Qwen4 Max
CritPt v2, hard subset (Artificial Analysis run)29.5%43.0%Qwen4 Max
GPQA Diamond (Epoch AI run)92.7%Not measuredOnly Qwen3.8 Max (0902) measured
Chess puzzles46.0%Not measuredOnly Qwen3.8 Max (0902) measured
Mystery game puzzles38.0%Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: hard prompts1500Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: expert1510Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: hard prompts in English1505Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Vision: diagrams1320Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Vision: homework1333Not measuredOnly Qwen3.8 Max (0902) measured
Neither model has a result on the other 10 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
Math benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
OTIS Mock AIME100.0%Not measuredOnly Qwen3.8 Max (0902) measured
FrontierMath, tiers 1 to 374.7%Not measuredOnly Qwen3.8 Max (0902) measured
FrontierMath, tier 446.3%Not measuredOnly Qwen3.8 Max (0902) measured
ProofBench58.0%Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: math1505Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: mathematical work1495Not measuredOnly Qwen3.8 Max (0902) measured
MathArena47.0%Not measuredOnly Qwen3.8 Max (0902) measured
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index01
Knowledge benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
MMLU Pro88.6%Not measuredOnly Qwen3.8 Max (0902) measured
LegalBench (Vals AI)83.6%Not measuredOnly Qwen3.8 Max (0902) measured
CorpFin (Vals AI)65.8%Not measuredOnly Qwen3.8 Max (0902) measured
TaxEval (Vals AI)75.5%Not measuredOnly Qwen3.8 Max (0902) measured
MortgageTax (Vals AI)64.0%Not measuredOnly Qwen3.8 Max (0902) measured
MedCode (Vals AI)40.7%Not measuredOnly Qwen3.8 Max (0902) measured
AA-Omniscience10.527.7Qwen4 Max
SimpleQA Verified47.3%Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: science1513Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: medicine and healthcare1512Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: legal and government1500Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: business and finance1481Not measuredOnly Qwen3.8 Max (0902) measured
FACTS Benchmark Suite60.3%Not measuredOnly Qwen3.8 Max (0902) measured
Neither model has a result on the other 8 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index01
Web design benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
Design Arena: websites1264Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: UI components1325Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: game dev1295Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: data viz1279Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: 3D1306Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: all code categories1279Not measuredOnly Qwen3.8 Max (0902) measured
Design Arena: ASCII art1184Not measuredOnly Qwen3.8 Max (0902) measured
LiveBench design76.6%79.1%Qwen4 Max
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index03
General benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
Vals Index59.1%63.2%Qwen4 Max
LiveBench data analysis78.4%83.4%Qwen4 Max
Public Benefits Bench (Vals AI)67.1%Not measuredOnly Qwen3.8 Max (0902) measured
SAGE (Vals AI)81.8%Not measuredOnly Qwen3.8 Max (0902) measured
MedScribe (Vals AI)85.0%Not measuredOnly Qwen3.8 Max (0902) measured
Vals Multimodal Index65.4%Not measuredOnly Qwen3.8 Max (0902) measured
Long context reasoning (AA-LCR)80.3%Not measuredOnly Qwen3.8 Max (0902) measured
GDPval-AA16741809Qwen4 Max
DTBench92.0%Not measuredOnly Qwen3.8 Max (0902) measured
LMCA46.2%Not measuredOnly Qwen3.8 Max (0902) measured
Furniture assembly20.0%Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text1485Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: instruction following1484Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: creative writing1484Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Vision1307Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: English1494Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: non-English1475Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: Chinese1544Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: German1474Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: Japanese1467Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: Korean1458Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: French1506Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: Spanish1492Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: Russian1487Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: Polish1482Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: multi-turn1492Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: longer queries1494Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: writing and language1480Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Text: entertainment and media1472Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Vision: English1306Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Vision: Chinese1354Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Vision: reading text1319Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Vision: humor1305Not measuredOnly Qwen3.8 Max (0902) measured
LMArena Vision: creative writing1301Not measuredOnly Qwen3.8 Max (0902) measured
Neither model has a result on the other 15 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Qwen3.8 Max (0902) and Qwen4 Max
BenchmarkQwen3.8 Max (0902)Qwen4 MaxLeader
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

Qwen3.8 Max (0902)$0.0050Fits in one request
Qwen4 Max$0.0036Fits in one request

Qwen4 Max costs less

Repository review

50K input + 3K output tokens

Qwen3.8 Max (0902)$0.118Fits in one request
Qwen4 Max$0.084Fits in one request

Qwen4 Max costs less

Agent loop

200K input + 10K output tokens

Qwen3.8 Max (0902)$0.460Fits in one request
Qwen4 Max$0.328Fits in one request

Qwen4 Max 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 Qwen3.8 Max (0902) and Qwen4 Max
SpecificationQwen3.8 Max (0902)Qwen4 Max
MakerQwenQwen
WeightsNo public weights listedNo public weights listed
ReleasedSeptember 3, 2026November 16, 2026
Context window1M1M
Input $/Mtok$2.00$1.425
Output $/Mtok$6.00$4.275
Coding index67.274.0
Independent measurements behind it54
Cost per solved task$1.32Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Qwen3.8 Max (0902) better than Qwen4 Max for coding?

Pick Qwen4 Max for coding: it scores 74.0 on the Vibe Coding Index against 67.2 for Qwen3.8 Max (0902), though on 4 benchmarks against 5. Pick Qwen3.8 Max (0902) for work measured by Tax Agent Bench (Vals AI) or Legal Research Bench (Vals AI), where it leads.

Which is cheaper, Qwen3.8 Max (0902) or Qwen4 Max?

Qwen3.8 Max (0902) costs $2.00 per million input tokens and $6.00 per million output tokens; Qwen4 Max costs $1.425 and $4.275. Cost per solved task is known for Qwen3.8 Max (0902) only.

Which has the bigger context window, Qwen3.8 Max (0902) or Qwen4 Max?

Neither: both take 1M tokens.

How many benchmarks is the Qwen3.8 Max (0902) vs Qwen4 Max comparison based on?

26 benchmarks have a result for both models as of November 22, 2026. Qwen3.8 Max (0902)'s coding index rests on 5 independent benchmarks and Qwen4 Max's on 4. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/qwen3-8-max-0902-vs-qwen4-max" title="Qwen3.8 Max (0902) vs Qwen4 Max, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/qwen3-8-max-0902-vs-qwen4-max">LLMBenchmark.ai, Qwen3.8 Max (0902) vs Qwen4 Max</a></p>