Enter opens the highlighted result. Esc closes.

GLM 5.3 Flash vs Qwen4 Max

Qwen4 Max leads GLM 5.3 Flash on coding, 74.0 to 56.3 on the Vibe Coding Index, and GLM 5.3 Flash costs 8.6x less per output token ($0.50 vs $4.275 per million). Updated November 22, 2026; 26 benchmarks have a result for both.

GLM 5.3 FlashZ.ai
Coding
56.3
Overall
60.8
Output $/M
$0.50
2
Qwen4 MaxQwen
Coding
74.0
Overall
66.6
Output $/M
$4.275
24
vs boards won of 26 shared

GLM 5.3 Flash or Qwen4 Max: which to pick

Pick Qwen4 Max for coding: it scores 74.0 on the Vibe Coding Index against 56.3 for GLM 5.3 Flash, though on 4 benchmarks against 6.

Pick GLM 5.3 Flash for output tokens 8.6x cheaper ($0.50 vs $4.275 per million) and a larger context window (1.05M against 1M tokens).

Shared results
26
GLM 5.3 Flash only
86
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
GLM 5.3 Flash 56.3 from 6 benchmarks, Qwen4 Max 74.0 from 4 benchmarks.
Widest category
Reasoning, 19.9 points apart on average over 3 shared benchmarks.
Widest single gap
Vibe Code Bench (Vals AI): GLM 5.3 Flash 30.8% against Qwen4 Max 86.7%, 55.9 points apart on a 0 to 100 scale.
Output price
GLM 5.3 Flash $0.50 against Qwen4 Max $4.275 per million output tokens, 8.6x apart.
Context window
GLM 5.3 Flash takes 1.05M tokens, Qwen4 Max 1M.

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
CategoryGLM 5.3 FlashBarsQwen4 MaxTies
Coding110
Intelligence01
Agentic work15
Reasoning03
Knowledge01
Web design01
General03
All shared boards224
CodingLike-for-likeSix of these boards feed the Vibe Coding Index110
Coding benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 MaxLeader
LiveBench coding70.9%69.8%GLM 5.3 Flash
LMArena WebDev16101751Qwen4 Max
LiveCodeBench (Vals AI)80.5%Not measuredOnly GLM 5.3 Flash measured
SWE-bench Verified (Vals AI)92.0%Not measuredOnly GLM 5.3 Flash measured
Terminal-Bench 4 (Vals AI)37.4%48.5%Qwen4 Max
Vibe Code Bench (Vals AI)30.8%86.7%Qwen4 Max
Code migration (Vals AI)20.5%Not measuredOnly GLM 5.3 Flash measured
ProgramBench (Vals AI)0.0%Not measuredOnly GLM 5.3 Flash measured
IOI (Vals AI)52.5%Not measuredOnly GLM 5.3 Flash measured
Terminal-Bench 2.1 (Vals AI)62.9%Not measuredOnly GLM 5.3 Flash measured
Vibe Code Bench 1-100 (Vals AI)16.0%Not measuredOnly GLM 5.3 Flash measured
SkillsBench (Vals AI)40.2%Not measuredOnly GLM 5.3 Flash measured
CyberBench (Vals AI)67.9%Not measuredOnly GLM 5.3 Flash measured
SciCode51.6%Not measuredOnly GLM 5.3 Flash measured
Terminal-Bench 4.0 (Artificial Analysis run)32.8%48.5%Qwen4 Max
FrontierCode 2.017.6%26.5%Qwen4 Max
FrontierCode31.8%Not measuredOnly GLM 5.3 Flash measured
FrontierSWE18.1%Not measuredOnly GLM 5.3 Flash measured
ALE-Bench303.61831.0Qwen4 Max
LMArena Text: coding1509Not measuredOnly GLM 5.3 Flash measured
LMArena Text: software and IT1501Not measuredOnly GLM 5.3 Flash measured
LMArena WebDev: React16301785Qwen4 Max
LMArena WebDev: HTML15791752Qwen4 Max
LMArena WebDev: from an image15811675Qwen4 Max
GameBench 260.682.4Qwen4 Max
Terminal-Bench 4.036.7%Not measuredOnly GLM 5.3 Flash measured
Neither model has a result on the other 26 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 MaxLeader
AA Intelligence Index27.842.9Qwen4 Max
Epoch Capabilities Index153.8Not measuredOnly GLM 5.3 Flash measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index15
Agentic work benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 MaxLeader
LiveBench research73.9%84.7%Qwen4 Max
LiveBench browser59.8%80.7%Qwen4 Max
LiveBench long agentic20.7%46.6%Qwen4 Max
SREBench (Vals AI)0.8%Not measuredOnly GLM 5.3 Flash measured
Finance Agent (Vals AI)57.9%Not measuredOnly GLM 5.3 Flash measured
Tax Agent Bench (Vals AI)31.6%40.4%Qwen4 Max
Legal Research Bench (Vals AI)45.2%41.4%GLM 5.3 Flash
Harvey's Legal Agent Benchmark (Vals AI)10.4%21.3%Qwen4 Max
Excel Modeling Benchmark (Vals AI)55.9%Not measuredOnly GLM 5.3 Flash measured
Terminal-Bench-Science (Vals AI)2.9%Not measuredOnly GLM 5.3 Flash measured
APEX-Agents52.8%Not measuredOnly GLM 5.3 Flash measured
PostTrainBench36.8%Not measuredOnly GLM 5.3 Flash measured
Toolathlon-Verified78.4%Not measuredOnly GLM 5.3 Flash measured
Neither model has a result on the other 31 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index03
Reasoning benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 MaxLeader
GPQA Diamond86.4%Not measuredOnly GLM 5.3 Flash measured
LiveBench reasoning77.6%92.2%Qwen4 Max
MMMU Pro (Vals AI)86.0%Not measuredOnly GLM 5.3 Flash measured
GPQA Diamond (Artificial Analysis run)91.2%Not measuredOnly GLM 5.3 Flash measured
Humanity's Last Exam (Artificial Analysis run)39.9%Not measuredOnly GLM 5.3 Flash measured
HLE-Diamond (Artificial Analysis run)18.2%44.9%Qwen4 Max
CritPt v2, hard subset (Artificial Analysis run)24.5%43.0%Qwen4 Max
GPQA Diamond (Epoch AI run)90.2%Not measuredOnly GLM 5.3 Flash measured
ARC-AGI91.0%Not measuredOnly GLM 5.3 Flash measured
ARC-AGI-265.8%Not measuredOnly GLM 5.3 Flash measured
Chess puzzles21.0%Not measuredOnly GLM 5.3 Flash measured
Mystery game puzzles8.0%Not measuredOnly GLM 5.3 Flash measured
LMArena Text: hard prompts1490Not measuredOnly GLM 5.3 Flash measured
LMArena Text: expert1513Not measuredOnly GLM 5.3 Flash measured
LMArena Text: hard prompts in English1491Not measuredOnly GLM 5.3 Flash measured
LMArena Vision: diagrams1320Not measuredOnly GLM 5.3 Flash measured
LMArena Vision: homework1325Not measuredOnly GLM 5.3 Flash measured
Neither model has a result on the other 9 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
Math benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 MaxLeader
OTIS Mock AIME93.9%Not measuredOnly GLM 5.3 Flash measured
FrontierMath, tiers 1 to 355.8%Not measuredOnly GLM 5.3 Flash measured
FrontierMath, tier 417.1%Not measuredOnly GLM 5.3 Flash measured
ProofBench21.0%Not measuredOnly GLM 5.3 Flash measured
LMArena Text: math1508Not measuredOnly GLM 5.3 Flash measured
LMArena Text: mathematical work1497Not measuredOnly GLM 5.3 Flash measured
Neither model has a result on the other 8 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index01
Knowledge benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 MaxLeader
MMLU Pro86.1%Not measuredOnly GLM 5.3 Flash measured
LegalBench (Vals AI)83.9%Not measuredOnly GLM 5.3 Flash measured
TaxEval (Vals AI)75.6%Not measuredOnly GLM 5.3 Flash measured
AA-Omniscience0.727.7Qwen4 Max
LMArena Text: science1488Not measuredOnly GLM 5.3 Flash measured
LMArena Text: medicine and healthcare1476Not measuredOnly GLM 5.3 Flash measured
LMArena Text: legal and government1479Not measuredOnly GLM 5.3 Flash measured
LMArena Text: business and finance1461Not measuredOnly GLM 5.3 Flash measured
Neither model has a result on the other 13 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index01
Web design benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 MaxLeader
Design Arena: websites1252Not measuredOnly GLM 5.3 Flash measured
Design Arena: UI components1289Not measuredOnly GLM 5.3 Flash measured
Design Arena: game dev1253Not measuredOnly GLM 5.3 Flash measured
Design Arena: data viz1254Not measuredOnly GLM 5.3 Flash measured
Design Arena: 3D1281Not measuredOnly GLM 5.3 Flash measured
Design Arena: all code categories1259Not measuredOnly GLM 5.3 Flash measured
Design Arena: SVG1265Not measuredOnly GLM 5.3 Flash measured
Design Arena: ASCII art1261Not measuredOnly GLM 5.3 Flash measured
LiveBench design73.7%79.1%Qwen4 Max
GeneralLike-for-likePublic boards, shown as published, kept out of the index03
General benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 MaxLeader
Vals Index50.9%63.2%Qwen4 Max
LiveBench data analysis76.4%83.4%Qwen4 Max
SAGE (Vals AI)89.3%Not measuredOnly GLM 5.3 Flash measured
MedScribe (Vals AI)88.9%Not measuredOnly GLM 5.3 Flash measured
Long context reasoning (AA-LCR)80.0%Not measuredOnly GLM 5.3 Flash measured
GDPval-AA16481809Qwen4 Max
Surface Evolver Bench52.5%Not measuredOnly GLM 5.3 Flash measured
LMArena Text1471Not measuredOnly GLM 5.3 Flash measured
LMArena Text: instruction following1480Not measuredOnly GLM 5.3 Flash measured
LMArena Text: creative writing1442Not measuredOnly GLM 5.3 Flash measured
LMArena Vision1296Not measuredOnly GLM 5.3 Flash measured
LMArena Text: English1477Not measuredOnly GLM 5.3 Flash measured
LMArena Text: non-English1461Not measuredOnly GLM 5.3 Flash measured
LMArena Text: Chinese1527Not measuredOnly GLM 5.3 Flash measured
LMArena Text: German1457Not measuredOnly GLM 5.3 Flash measured
LMArena Text: Japanese1450Not measuredOnly GLM 5.3 Flash measured
LMArena Text: Korean1427Not measuredOnly GLM 5.3 Flash measured
LMArena Text: French1487Not measuredOnly GLM 5.3 Flash measured
LMArena Text: Spanish1478Not measuredOnly GLM 5.3 Flash measured
LMArena Text: Russian1469Not measuredOnly GLM 5.3 Flash measured
LMArena Text: Polish1451Not measuredOnly GLM 5.3 Flash measured
LMArena Text: multi-turn1474Not measuredOnly GLM 5.3 Flash measured
LMArena Text: longer queries1482Not measuredOnly GLM 5.3 Flash measured
LMArena Text: writing and language1457Not measuredOnly GLM 5.3 Flash measured
LMArena Text: entertainment and media1442Not measuredOnly GLM 5.3 Flash measured
LMArena Vision: English1300Not measuredOnly GLM 5.3 Flash measured
LMArena Vision: Chinese1357Not measuredOnly GLM 5.3 Flash measured
LMArena Vision: reading text1306Not measuredOnly GLM 5.3 Flash measured
LMArena Vision: humor1291Not measuredOnly GLM 5.3 Flash measured
LMArena Vision: recognizing things1275Not measuredOnly GLM 5.3 Flash measured
LMArena Vision: creative writing1272Not measuredOnly GLM 5.3 Flash measured
Neither model has a result on the other 18 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GLM 5.3 Flash and Qwen4 Max
BenchmarkGLM 5.3 FlashQwen4 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

GLM 5.3 Flash$0.0004Fits in one request
Qwen4 Max$0.0036Fits in one request

GLM 5.3 Flash costs less

Repository review

50K input + 3K output tokens

GLM 5.3 Flash$0.0090Fits in one request
Qwen4 Max$0.084Fits in one request

GLM 5.3 Flash costs less

Agent loop

200K input + 10K output tokens

GLM 5.3 Flash$0.035Fits in one request
Qwen4 Max$0.328Fits in one request

GLM 5.3 Flash 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 GLM 5.3 Flash and Qwen4 Max
SpecificationZ.ai GLM 5.3 FlashQwen4 Max
MakerZ.aiQwen
WeightsOpen weightsNo public weights listed
ReleasedAugust 26, 2026November 16, 2026
Context window1.05M1M
Input $/Mtok$0.15$1.425
Output $/Mtok$0.50$4.275
Coding index56.374.0
Independent measurements behind it64
Cost per solved task$0.02Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GLM 5.3 Flash better than Qwen4 Max for coding?

Pick Qwen4 Max for coding: it scores 74.0 on the Vibe Coding Index against 56.3 for GLM 5.3 Flash, though on 4 benchmarks against 6. Pick GLM 5.3 Flash for output tokens 8.6x cheaper ($0.50 vs $4.275 per million) and a larger context window (1.05M against 1M tokens).

Which is cheaper, GLM 5.3 Flash or Qwen4 Max?

GLM 5.3 Flash costs $0.15 per million input tokens and $0.50 per million output tokens; Qwen4 Max costs $1.425 and $4.275. Cost per solved task is known for GLM 5.3 Flash only.

Which has the bigger context window, GLM 5.3 Flash or Qwen4 Max?

GLM 5.3 Flash, with 1.05M tokens against 1M.

How many benchmarks is the GLM 5.3 Flash vs Qwen4 Max comparison based on?

26 benchmarks have a result for both models as of November 22, 2026. GLM 5.3 Flash's coding index rests on 6 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/glm-5-3-flash-vs-qwen4-max" title="GLM 5.3 Flash 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/glm-5-3-flash-vs-qwen4-max">LLMBenchmark.ai, GLM 5.3 Flash vs Qwen4 Max</a></p>