GLM 4.5V vs Qwen3 VL 30B A3B Instruct
Qwen3 VL 30B A3B Instruct leads GLM 4.5V on the Artificial Analysis Intelligence Index, 7.9 to 7.6, and Qwen3 VL 30B A3B Instruct costs 2.6x less per output token ($0.70 vs $1.80 per million). Updated October 22, 2026; 10 benchmarks have a result for both.
GLM 4.5V or Qwen3 VL 30B A3B Instruct: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick GLM 4.5V for work measured by LiveCodeBench (Artificial Analysis run) or AA Intelligence Index, where it leads.
- Shared results
- 10
- GLM 4.5V only
- 31
- Qwen3 VL 30B A3B Instruct 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.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Qwen3 VL 30B A3B Instruct
Qwen3 VL 30B A3B Instruct leads, 7.9 against 7.6.
Same board, both measured
Long documents
Listed context window
Qwen3 VL 30B A3B Instruct
Qwen3 VL 30B A3B Instruct takes 262K tokens against 66K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3 VL 30B A3B Instruct
GLM 4.5V $0.0015, Qwen3 VL 30B A3B Instruct $0.0006 per turn.
List prices
At a glance
- Coding index
- GLM 4.5V not benchmarked, Qwen3 VL 30B A3B Instruct not benchmarked.
- Widest category
- General, 10.9 points apart on average over 1 shared benchmark.
- Widest single gap
- IFBench: GLM 4.5V 34.2% against Qwen3 VL 30B A3B Instruct 45.1%, 10.9 points apart on a 0 to 100 scale.
- Output price
- GLM 4.5V $1.80 against Qwen3 VL 30B A3B Instruct $0.70 per million output tokens, 2.6x apart.
- Context window
- GLM 4.5V takes 66K tokens, Qwen3 VL 30B A3B Instruct 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.
| Category | GLM 4.5V | Bars | Qwen3 VL 30B A3B Instruct | Ties |
|---|---|---|---|---|
| Coding | 1 | 1 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 0 | 2 | ||
| Math | 0 | 2 | ||
| Knowledge | 0 | 1 | ||
| General | 0 | 1 | ||
| All shared boards | 2 | 8 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index11
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | Leader |
|---|---|---|---|
| LiveCodeBench (Artificial Analysis run) | 60.4% | 69.7% | Qwen3 VL 30B A3B Instruct |
| Terminal-Bench Hard (Artificial Analysis run) | 6.8% | 6.1% | GLM 4.5V |
| LMArena Text: coding | 1347 | Not measured | Only GLM 4.5V measured |
| LMArena Text: software and IT | 1353 | Not measured | Only GLM 4.5V measured |
| Neither model has a result on the other 35 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | 7.6 | 7.9 | Qwen3 VL 30B A3B Instruct |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 22.5% | 19.9% | GLM 4.5V |
| Neither model has a result on the other 40 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 68.4% | 72.0% | Qwen3 VL 30B A3B Instruct |
| Humanity's Last Exam (Artificial Analysis run) | 6.3% | 8.9% | Qwen3 VL 30B A3B Instruct |
| LMArena Text: hard prompts | 1335 | Not measured | Only GLM 4.5V measured |
| LMArena Text: expert | 1353 | Not measured | Only GLM 4.5V measured |
| LMArena Text: hard prompts in English | 1357 | Not measured | Only GLM 4.5V measured |
| LMArena Vision: diagrams | 1151 | Not measured | Only GLM 4.5V measured |
| LMArena Vision: homework | 1163 | Not measured | Only GLM 4.5V measured |
| Kagi LLM Benchmark | 59.8% | Not measured | Only GLM 4.5V measured |
| Neither model has a result on the other 14 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | Leader |
|---|---|---|---|
| AA Math Index | 73.0 | 82.3 | Qwen3 VL 30B A3B Instruct |
| AIME 2025 (Artificial Analysis run) | 73.0% | 82.3% | Qwen3 VL 30B A3B Instruct |
| LMArena Text: math | 1354 | Not measured | Only GLM 4.5V measured |
| LMArena Text: mathematical work | 1336 | Not measured | Only GLM 4.5V measured |
| Neither model has a result on the other 9 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 78.8% | 80.7% | Qwen3 VL 30B A3B Instruct |
| LMArena Text: science | 1330 | Not measured | Only GLM 4.5V measured |
| LMArena Text: medicine and healthcare | 1349 | Not measured | Only GLM 4.5V measured |
| LMArena Text: legal and government | 1338 | Not measured | Only GLM 4.5V measured |
| LMArena Text: business and finance | 1309 | Not measured | Only GLM 4.5V measured |
| Neither model has a result on the other 13 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | Leader |
|---|---|---|---|
| IFBench | 34.2% | 45.1% | Qwen3 VL 30B A3B Instruct |
| Long context reasoning (AA-LCR) | 0.0% | Not measured | Only GLM 4.5V measured |
| LMArena Text | 1333 | Not measured | Only GLM 4.5V measured |
| LMArena Text: instruction following | 1311 | Not measured | Only GLM 4.5V measured |
| LMArena Text: creative writing | 1294 | Not measured | Only GLM 4.5V measured |
| LMArena Vision | 1154 | Not measured | Only GLM 4.5V measured |
| LMArena Text: English | 1362 | Not measured | Only GLM 4.5V measured |
| LMArena Text: non-English | 1303 | Not measured | Only GLM 4.5V measured |
| LMArena Text: Chinese | 1337 | Not measured | Only GLM 4.5V measured |
| LMArena Text: Spanish | 1336 | Not measured | Only GLM 4.5V measured |
| LMArena Text: Russian | 1298 | Not measured | Only GLM 4.5V measured |
| LMArena Text: Polish | 1302 | Not measured | Only GLM 4.5V measured |
| LMArena Text: multi-turn | 1333 | Not measured | Only GLM 4.5V measured |
| LMArena Text: longer queries | 1305 | Not measured | Only GLM 4.5V measured |
| LMArena Text: writing and language | 1302 | Not measured | Only GLM 4.5V measured |
| LMArena Text: entertainment and media | 1303 | Not measured | Only GLM 4.5V measured |
| LMArena Vision: English | 1183 | Not measured | Only GLM 4.5V measured |
| LMArena Vision: reading text | 1151 | Not measured | Only GLM 4.5V measured |
| Neither model has a result on the other 28 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GLM 4.5V | Qwen3 VL 30B A3B Instruct | 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 VL 30B A3B Instruct costs less
Repository review
50K input + 3K output tokens
Qwen3 VL 30B A3B Instruct costs less
Agent loop
200K input + 10K output tokens
Qwen3 VL 30B A3B Instruct 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 4.5V | Qwen3 VL 30B A3B Instruct |
|---|---|---|
| Maker | Z.ai | Qwen |
| Weights | Open weights | Open weights |
| Released | August 11, 2025 | October 6, 2025 |
| Context window | 66K | 262K |
| Input $/Mtok | $0.60 | $0.20 |
| Output $/Mtok | $1.80 | $0.70 |
| 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 4.5V better than Qwen3 VL 30B A3B Instruct for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GLM 4.5V for work measured by LiveCodeBench (Artificial Analysis run) or AA Intelligence Index, where it leads.
Which is cheaper, GLM 4.5V or Qwen3 VL 30B A3B Instruct?
GLM 4.5V costs $0.60 per million input tokens and $1.80 per million output tokens; Qwen3 VL 30B A3B Instruct costs $0.20 and $0.70. Cost per solved task is known for neither of them.
Which has the bigger context window, GLM 4.5V or Qwen3 VL 30B A3B Instruct?
Qwen3 VL 30B A3B Instruct, with 262K tokens against 66K.
How many benchmarks is the GLM 4.5V vs Qwen3 VL 30B A3B Instruct comparison based on?
10 benchmarks have a result for both models as of October 22, 2026. GLM 4.5V's coding index rests on 0 independent benchmarks and Qwen3 VL 30B A3B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.