Enter opens the highlighted result. Esc closes.

GLM 4.5 vs Qwen3 30B A3B Instruct 2507

GLM 4.5 leads Qwen3 30B A3B Instruct 2507 on the Artificial Analysis Intelligence Index, 12.8 to 7.5, and Qwen3 30B A3B Instruct 2507 costs 7.3x less per output token ($0.30 vs $2.20 per million). Updated October 22, 2026; 39 benchmarks have a result for both.

Replace GLM 4.5 Replace Qwen3 30B A3B Instruct 2507
GLM 4.5Z.ai
Coding
No data
Overall
No data
Output $/M
$2.20
36
Qwen3 30B A3B Instruct 2507Qwen
Coding
No data
Overall
41.5
Output $/M
$0.30
3
vs boards won of 39 shared

GLM 4.5 or Qwen3 30B A3B Instruct 2507: which to pick

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

Pick GLM 4.5 for work measured by IFBench or Long context reasoning (AA-LCR), where it leads.

Model AGLM 4.5Z.aiNot benchmarkedNo coding benchmarkNo cost per solved taskModel BQwen3 30B A3B Instruct 2507QwenNot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
39
GLM 4.5 only
22
Qwen3 30B A3B Instruct 2507 only
6
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 4.5 not benchmarked, Qwen3 30B A3B Instruct 2507 not benchmarked.
Widest category
Agentic work, 14.9 points apart on average over 1 shared benchmark.
Widest single gap
Terminal-Bench Hard (Artificial Analysis run): GLM 4.5 22.0% against Qwen3 30B A3B Instruct 2507 6.1%, 15.9 points apart on a 0 to 100 scale.
Output price
GLM 4.5 $2.20 against Qwen3 30B A3B Instruct 2507 $0.30 per million output tokens, 7.3x apart.
Context window
GLM 4.5 takes 131K tokens, Qwen3 30B A3B Instruct 2507 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
CategoryGLM 4.5BarsQwen3 30B A3B Instruct 2507Ties
Coding40
Intelligence10
Agentic work10
Reasoning50
Math40
Knowledge50
General163
All shared boards363
CodingLike-for-likeSix of these boards feed the Vibe Coding Index40
Coding benchmarks for GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
SWE-bench Verified64.2%Not measuredOnly GLM 4.5 measured
LiveCodeBench (Vals AI)67.5%Not measuredOnly GLM 4.5 measured
LiveCodeBench (Artificial Analysis run)73.8%70.7%GLM 4.5
SciCodeNot measured33.0%Only Qwen3 30B A3B Instruct 2507 measured
Terminal-Bench Hard (Artificial Analysis run)22.0%6.1%GLM 4.5
WeirdML40.6%Not measuredOnly GLM 4.5 measured
ALE-Bench344.8Not measuredOnly GLM 4.5 measured
LMArena Text: coding14341416GLM 4.5
LMArena Text: software and IT14401417GLM 4.5
Neither model has a result on the other 30 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
Intelligence benchmarks for GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
AA Intelligence Index12.87.5GLM 4.5
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 GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
τ²-Bench Telecom43.0%28.1%GLM 4.5
Berkeley Function Calling Leaderboard (BFCL)Not measured41.4%Only Qwen3 30B A3B Instruct 2507 measured
Neither model has a result on the other 39 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index50
Reasoning benchmarks for GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
GPQA Diamond72.2%Not measuredOnly GLM 4.5 measured
GPQA Diamond (Artificial Analysis run)78.2%70.7%GLM 4.5
Humanity's Last Exam (Artificial Analysis run)13.0%10.3%GLM 4.5
GPQA Diamond (Epoch AI run)Not measured55.6%Only Qwen3 30B A3B Instruct 2507 measured
LMArena Text: hard prompts14291398GLM 4.5
LMArena Text: expert14331397GLM 4.5
LMArena Text: hard prompts in English14291406GLM 4.5
Kagi LLM Benchmark57.9%Not measuredOnly GLM 4.5 measured
Neither model has a result on the other 14 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index40
Math benchmarks for GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
AIME86.7%Not measuredOnly GLM 4.5 measured
MATH 500 (Vals AI)94.0%Not measuredOnly GLM 4.5 measured
MGSM (Vals AI)90.8%Not measuredOnly GLM 4.5 measured
AA Math Index73.766.3GLM 4.5
AIME 2025 (Artificial Analysis run)73.7%66.3%GLM 4.5
OTIS Mock AIMENot measured62.2%Only Qwen3 30B A3B Instruct 2507 measured
LMArena Text: math14271394GLM 4.5
LMArena Text: mathematical work14301411GLM 4.5
Neither model has a result on the other 5 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index50
Knowledge benchmarks for GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
MMLU Pro81.2%Not measuredOnly GLM 4.5 measured
LegalBench (Vals AI)75.6%Not measuredOnly GLM 4.5 measured
CorpFin (Vals AI)61.0%Not measuredOnly GLM 4.5 measured
TaxEval (Vals AI)72.4%Not measuredOnly GLM 4.5 measured
MedQA (Vals AI)90.0%Not measuredOnly GLM 4.5 measured
MMLU Pro (Artificial Analysis run)83.5%80.5%GLM 4.5
LMArena Text: science14441393GLM 4.5
LMArena Text: medicine and healthcare14551392GLM 4.5
LMArena Text: legal and government14381387GLM 4.5
LMArena Text: business and finance14201389GLM 4.5
Neither model has a result on the other 8 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
Web design benchmarks for GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
Design Arena: websites1169Not measuredOnly GLM 4.5 measured
Design Arena: UI components1141Not measuredOnly GLM 4.5 measured
Design Arena: game dev1144Not measuredOnly GLM 4.5 measured
Design Arena: data viz1159Not measuredOnly GLM 4.5 measured
Design Arena: 3D1170Not measuredOnly GLM 4.5 measured
Design Arena: all code categories1165Not measuredOnly GLM 4.5 measured
Design Arena: SVG1096Not measuredOnly GLM 4.5 measured
Neither model has a result on the other 2 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index163
General benchmarks for GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
IFBench44.1%50.7%Qwen3 30B A3B Instruct 2507
Long context reasoning (AA-LCR)52.7%61.3%Qwen3 30B A3B Instruct 2507
DTBenchNot measured67.2%Only Qwen3 30B A3B Instruct 2507 measured
LMCANot measured22.4%Only Qwen3 30B A3B Instruct 2507 measured
Creative writing (Lech Mazur)73.4%Not measuredOnly GLM 4.5 measured
LMArena Text14301384GLM 4.5
LMArena Text: instruction following14041363GLM 4.5
LMArena Text: creative writing13951317GLM 4.5
LMArena Text: English14371391GLM 4.5
LMArena Text: non-English14171372GLM 4.5
LMArena Text: Chinese14651432GLM 4.5
LMArena Text: German14071379GLM 4.5
LMArena Text: Japanese14151337GLM 4.5
LMArena Text: Korean13801330GLM 4.5
LMArena Text: French14181419Qwen3 30B A3B Instruct 2507
LMArena Text: Spanish14541404GLM 4.5
LMArena Text: Russian14131369GLM 4.5
LMArena Text: Polish14221353GLM 4.5
LMArena Text: multi-turn14151378GLM 4.5
LMArena Text: longer queries14121378GLM 4.5
LMArena Text: writing and language14031336GLM 4.5
LMArena Text: entertainment and media14041334GLM 4.5
Neither model has a result on the other 24 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GLM 4.5 and Qwen3 30B A3B Instruct 2507
BenchmarkGLM 4.5Qwen3 30B A3B Instruct 2507Leader
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 4.5$0.0017Fits in one request
Qwen3 30B A3B Instruct 2507$0.0002Fits in one request

Qwen3 30B A3B Instruct 2507 costs less

Repository review

50K input + 3K output tokens

GLM 4.5$0.037Fits in one request
Qwen3 30B A3B Instruct 2507$0.0054Fits in one request

Qwen3 30B A3B Instruct 2507 costs less

Agent loop

200K input + 10K output tokens

GLM 4.5$0.142Exceeds the context window
Qwen3 30B A3B Instruct 2507$0.021Fits in one request

Qwen3 30B A3B Instruct 2507 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 4.5 and Qwen3 30B A3B Instruct 2507
SpecificationZ.ai GLM 4.5Qwen3 30B A3B Instruct 2507
MakerZ.aiQwen
WeightsOpen weightsOpen weights
ReleasedJuly 25, 2025July 29, 2025
Context window131K262K
Input $/Mtok$0.60$0.09
Output $/Mtok$2.20$0.30
Coding indexNot benchmarkedNot benchmarked
Independent measurements behind it00
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GLM 4.5 better than Qwen3 30B A3B Instruct 2507 for coding?

Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GLM 4.5 for work measured by IFBench or Long context reasoning (AA-LCR), where it leads.

Which is cheaper, GLM 4.5 or Qwen3 30B A3B Instruct 2507?

GLM 4.5 costs $0.60 per million input tokens and $2.20 per million output tokens; Qwen3 30B A3B Instruct 2507 costs $0.09 and $0.30. Cost per solved task is known for neither of them.

Which has the bigger context window, GLM 4.5 or Qwen3 30B A3B Instruct 2507?

Qwen3 30B A3B Instruct 2507, with 262K tokens against 131K.

How many benchmarks is the GLM 4.5 vs Qwen3 30B A3B Instruct 2507 comparison based on?

39 benchmarks have a result for both models as of October 22, 2026. GLM 4.5's coding index rests on 0 independent benchmarks and Qwen3 30B A3B Instruct 2507's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next