# GLM-4.6V vs Qwen3-Coder 480B-A35B Instruct

> GLM-4.6V is the stronger model overall, scoring 40.9 to 37.7 on the Noometry Index.

- Canonical page: https://noometry.com/compare/glm-4-6v-vs-qwen3-coder-480b-a35b-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 11

## Summary

- They share 11 benchmarks with published results for both. GLM-4.6V scores higher in 5 categories and Qwen3-Coder 480B-A35B Instruct in 2 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in coding, where GLM-4.6V leads 40.2 to 36.2.
- GLM-4.6V is cheaper at $0.30 / $0.90 per million input/output tokens, against $1.50 / $7.50 for Qwen3-Coder 480B-A35B Instruct.
- Qwen3-Coder 480B-A35B Instruct accepts more context: 262K tokens versus 128K.

## Snapshot

| | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| Provider | Z.ai (Zhipu) | Alibaba (Qwen) |
| Noometry Index | 40.9 | 37.7 |
| Rank | 133 | 186 |
| Context | 128K | 262K |
| Input $/M | $0.30 | $1.50 |
| Output $/M | $0.90 | $7.50 |
| Weights | Open | Open |

## Coding

- GLM-4.6V: 40.2 (#128)
- Qwen3-Coder 480B-A35B Instruct: 36.2 (#199)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Coding | 1390 | 1413 |
| SWE-bench Verified (bash only) | — | 55.4% |
| LMArena WebDev | — | 1275 |
| GSO | — | 4.9% |
| WeirdML | — | 41.2% |
| ALE-Bench | — | 461.45 |
| AlgoTune | — | 1.44 |

## Agentic & Tool Use

- GLM-4.6V: —
- Qwen3-Coder 480B-A35B Instruct: 23.7 (#116)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| Terminal-Bench | — | 27.2% |

## Reasoning

- GLM-4.6V: 26.5 (#117)
- Qwen3-Coder 480B-A35B Instruct: 24.7 (#141)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1368 | 1372 |
| Kagi LLM Benchmark | — | 49.5% |

## Math

- GLM-4.6V: —
- Qwen3-Coder 480B-A35B Instruct: 34.9 (#151)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Math | — | 1364 |

## Knowledge

- GLM-4.6V: 37.4 (#148)
- Qwen3-Coder 480B-A35B Instruct: 36.4 (#159)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Expert | 1372 | 1337 |

## Multimodal

- GLM-4.6V: 34.2 (#83)
- Qwen3-Coder 480B-A35B Instruct: —

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Vision | 1164 | — |

## Multilingual

- GLM-4.6V: 48.8 (#135)
- Qwen3-Coder 480B-A35B Instruct: 47.8 (#143)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Non-English | 1359 | 1345 |
| LMArena Chinese | 1426 | 1356 |
| LMArena Russian | 1341 | 1367 |
| LMArena French | — | 1394 |
| LMArena German | — | 1324 |
| LMArena Japanese | — | 1309 |
| LMArena Korean | — | 1307 |
| LMArena Spanish | — | 1360 |

## Instruction Following

- GLM-4.6V: 71.6 (#145)
- Qwen3-Coder 480B-A35B Instruct: 71.8 (#142)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Instruction Following | 1351 | 1356 |

## Long Context

- GLM-4.6V: 41.3 (#138)
- Qwen3-Coder 480B-A35B Instruct: 42.0 (#126)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Longer Query | 1358 | 1379 |

## Writing & Preference

- GLM-4.6V: 56.8 (#132)
- Qwen3-Coder 480B-A35B Instruct: 55.5 (#142)

| Benchmark | GLM-4.6V | Qwen3-Coder 480B-A35B Instruct |
|---|---|---|
| LMArena Text | 1377 | 1356 |
| LMArena Creative Writing | 1347 | 1333 |
| LMArena Multi-Turn | 1359 | 1365 |

## FAQ

### Is GLM-4.6V better than Qwen3-Coder 480B-A35B Instruct?

GLM-4.6V is the stronger model overall, scoring 40.9 to 37.7 on the Noometry Index.

### Which is cheaper, GLM-4.6V or Qwen3-Coder 480B-A35B Instruct?

GLM-4.6V is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; Qwen3-Coder 480B-A35B Instruct lists at $1.50 and $7.50.

### Is GLM-4.6V or Qwen3-Coder 480B-A35B Instruct better for coding?

GLM-4.6V scores higher on coding benchmarks: 40.2 versus 36.2 in the Noometry coding category.

### Which has the bigger context window?

Qwen3-Coder 480B-A35B Instruct does, with 262K tokens against 128K.

### How many benchmarks do GLM-4.6V and Qwen3-Coder 480B-A35B Instruct share?

11 benchmarks have published results for both models. GLM-4.6V has 12 scored results on Noometry and Qwen3-Coder 480B-A35B Instruct has 25.