GLM 4.5 Air vs Qwen3 32B
GLM 4.5 Air leads Qwen3 32B on the Artificial Analysis Intelligence Index, 11.1 to 7.3, and Qwen3 32B costs 2.8x less per output token ($0.40 vs $1.10 per million). Updated October 22, 2026; 35 benchmarks have a result for both.
GLM 4.5 Air or Qwen3 32B: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick GLM 4.5 Air for work measured by Humanity's Last Exam (Artificial Analysis run) or Kagi LLM Benchmark, where it leads.
- Shared results
- 35
- GLM 4.5 Air only
- 13
- Qwen3 32B only
- 10
- 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.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
GLM 4.5 Air
GLM 4.5 Air leads, 11.1 against 7.3.
Same board, both measured
Web design
Design Arena: websites
No call
Only GLM 4.5 Air has a Design Arena: websites result, 1146.
Directional only
Long documents
Listed context window
No call
Both take 131K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3 32B
GLM 4.5 Air $0.0008, Qwen3 32B $0.0003 per turn.
List prices
At a glance
- Coding index
- GLM 4.5 Air not benchmarked, Qwen3 32B not benchmarked.
- Widest category
- Agentic work, 16.7 points apart on average over 1 shared benchmark.
- Widest single gap
- Long context reasoning (AA-LCR): GLM 4.5 Air 46.7% against Qwen3 32B 0.0%, 46.7 points apart on a 0 to 100 scale.
- Output price
- GLM 4.5 Air $1.10 against Qwen3 32B $0.40 per million output tokens, 2.8x apart.
- Context window
- Both take 131K 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.
| Category | GLM 4.5 Air | Bars | Qwen3 32B | Ties |
|---|---|---|---|---|
| Coding | 4 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 4 | 2 | ||
| Math | 2 | 2 | ||
| Knowledge | 5 | 0 | ||
| General | 14 | 0 | ||
| All shared boards | 31 | 4 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index40
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| Aider polyglot | Not measured | 40.0% | Only Qwen3 32B measured |
| LiveCodeBench (Artificial Analysis run) | 68.4% | 54.6% | GLM 4.5 Air |
| SciCode | Not measured | 36.0% | Only Qwen3 32B measured |
| Terminal-Bench Hard (Artificial Analysis run) | 20.5% | 3.0% | GLM 4.5 Air |
| GSO-Bench | 2.9% | Not measured | Only GLM 4.5 Air measured |
| LMArena Text: coding | 1397 | 1357 | GLM 4.5 Air |
| LMArena Text: software and IT | 1401 | 1366 | GLM 4.5 Air |
| Neither model has a result on the other 32 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| AA Intelligence Index | 11.1 | 7.3 | GLM 4.5 Air |
| Epoch Capabilities Index | Not measured | 138.4 | Only Qwen3 32B measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 46.5% | 29.8% | GLM 4.5 Air |
| Berkeley Function Calling Leaderboard (BFCL) | Not measured | 48.7% | Only Qwen3 32B measured |
| BrowseComp-Plus | Not measured | 10.7% | Only Qwen3 32B measured |
| Neither model has a result on the other 38 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index42
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 73.3% | 66.8% | GLM 4.5 Air |
| Humanity's Last Exam (Artificial Analysis run) | 7.0% | 7.4% | Qwen3 32B |
| GPQA Diamond (Epoch AI run) | Not measured | 65.7% | Only Qwen3 32B measured |
| LMArena Text: hard prompts | 1379 | 1334 | GLM 4.5 Air |
| LMArena Text: expert | 1370 | 1362 | GLM 4.5 Air |
| LMArena Text: hard prompts in English | 1388 | 1350 | GLM 4.5 Air |
| Kagi LLM Benchmark | 43.0% | 48.7% | Qwen3 32B |
| Neither model has a result on the other 15 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index22
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| AA Math Index | 80.7 | 73.0 | GLM 4.5 Air |
| AIME 2025 (Artificial Analysis run) | 80.7% | 73.0% | GLM 4.5 Air |
| OTIS Mock AIME | Not measured | 66.9% | Only Qwen3 32B measured |
| LMArena Text: math | 1396 | 1399 | Qwen3 32B |
| LMArena Text: mathematical work | 1396 | 1420 | Qwen3 32B |
| Neither model has a result on the other 8 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index50
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 81.5% | 79.8% | GLM 4.5 Air |
| LMArena Text: science | 1407 | 1372 | GLM 4.5 Air |
| LMArena Text: medicine and healthcare | 1402 | 1368 | GLM 4.5 Air |
| LMArena Text: legal and government | 1382 | 1324 | GLM 4.5 Air |
| LMArena Text: business and finance | 1371 | 1318 | GLM 4.5 Air |
| Neither model has a result on the other 13 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| Design Arena: websites | 1146 | Not measured | Only GLM 4.5 Air measured |
| Design Arena: UI components | 1120 | Not measured | Only GLM 4.5 Air measured |
| Design Arena: game dev | 1094 | Not measured | Only GLM 4.5 Air measured |
| Design Arena: data viz | 1186 | Not measured | Only GLM 4.5 Air measured |
| Design Arena: 3D | 1122 | Not measured | Only GLM 4.5 Air measured |
| Design Arena: all code categories | 1137 | Not measured | Only GLM 4.5 Air measured |
| Design Arena: SVG | 1071 | Not measured | Only GLM 4.5 Air measured |
| Neither model has a result on the other 2 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index140
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| IFBench | 37.6% | 36.3% | GLM 4.5 Air |
| Long context reasoning (AA-LCR) | 46.7% | 0.0% | GLM 4.5 Air |
| DTBench | Not measured | 67.5% | Only Qwen3 32B measured |
| LMCA | Not measured | 17.3% | Only Qwen3 32B measured |
| Fiction.LiveBench | Not measured | 74.2% | Only Qwen3 32B measured |
| LMArena Text | 1384 | 1340 | GLM 4.5 Air |
| LMArena Text: instruction following | 1354 | 1305 | GLM 4.5 Air |
| LMArena Text: creative writing | 1343 | 1297 | GLM 4.5 Air |
| LMArena Text: English | 1397 | 1359 | GLM 4.5 Air |
| LMArena Text: non-English | 1366 | 1317 | GLM 4.5 Air |
| LMArena Text: Chinese | 1426 | 1357 | GLM 4.5 Air |
| LMArena Text: German | 1377 | 1341 | GLM 4.5 Air |
| LMArena Text: Japanese | 1348 | Not measured | Only GLM 4.5 Air measured |
| LMArena Text: Korean | 1308 | Not measured | Only GLM 4.5 Air measured |
| LMArena Text: French | 1400 | Not measured | Only GLM 4.5 Air measured |
| LMArena Text: Spanish | 1385 | Not measured | Only GLM 4.5 Air measured |
| LMArena Text: Russian | 1373 | 1311 | GLM 4.5 Air |
| LMArena Text: Polish | 1368 | Not measured | Only GLM 4.5 Air measured |
| LMArena Text: multi-turn | 1371 | 1331 | GLM 4.5 Air |
| LMArena Text: longer queries | 1366 | 1327 | GLM 4.5 Air |
| LMArena Text: writing and language | 1354 | 1298 | GLM 4.5 Air |
| LMArena Text: entertainment and media | 1339 | 1279 | GLM 4.5 Air |
| 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
| Benchmark | GLM 4.5 Air | Qwen3 32B | Leader |
|---|---|---|---|
| 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 32B costs less
Repository review
50K input + 3K output tokens
Qwen3 32B costs less
Agent loop
200K input + 10K output tokens
Qwen3 32B 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.
| Specification | Z.ai GLM 4.5 Air | Qwen3 32B |
|---|---|---|
| Maker | Z.ai | Qwen |
| Weights | Open weights | Open weights |
| Released | July 25, 2025 | April 28, 2025 |
| Context window | 131K | 131K |
| Input $/Mtok | $0.20 | $0.14 |
| Output $/Mtok | $1.10 | $0.40 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is GLM 4.5 Air better than Qwen3 32B for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GLM 4.5 Air for work measured by Humanity's Last Exam (Artificial Analysis run) or Kagi LLM Benchmark, where it leads.
Which is cheaper, GLM 4.5 Air or Qwen3 32B?
GLM 4.5 Air costs $0.20 per million input tokens and $1.10 per million output tokens; Qwen3 32B costs $0.14 and $0.40. Cost per solved task is known for neither of them.
Which has the bigger context window, GLM 4.5 Air or Qwen3 32B?
Neither: both take 131K tokens.
How many benchmarks is the GLM 4.5 Air vs Qwen3 32B comparison based on?
35 benchmarks have a result for both models as of October 22, 2026. GLM 4.5 Air's coding index rests on 0 independent benchmarks and Qwen3 32B's on 0. Where only one model has a result, the page says so instead of naming a winner.