# GLM-4.5 vs MiniMax-M2

> GLM-4.5 is the stronger model overall, scoring 42.0 to 37.4 on the Noometry Index. MiniMax-M2 costs 1.9× less per token, which makes it the better buy when GLM-4.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/glm-4-5-vs-minimax-m2
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. GLM-4.5 scores higher in 6 categories and MiniMax-M2 in 2 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-4.5 leads 28.6 to 19.4.
- The biggest single-benchmark swing is SWE-bench Verified (bash only): 54.2% for GLM-4.5 and 61% for MiniMax-M2.
- MiniMax-M2 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.60 / $2.20 for GLM-4.5.
- MiniMax-M2 accepts more context: 205K tokens versus 131K.

## Snapshot

| | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| Provider | Z.ai (Zhipu) | MiniMax |
| Noometry Index | 42.0 | 37.4 |
| Rank | 122 | 204 |
| Context | 131K | 205K |
| Input $/M | $0.60 | $0.30 |
| Output $/M | $2.20 | $1.20 |
| Weights | Open | Open |

## Coding

- GLM-4.5: 41.4 (#125)
- MiniMax-M2: 39.3 (#159)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| SWE-bench Verified (bash only) | 54.2% | 61% |
| LMArena Coding | 1434 | 1370 |
| LMArena WebDev | — | 1297 |
| WeirdML | 40.6% | — |
| ALE-Bench | 344.82 | — |
| AlgoTune | 1.52 | — |

## Agentic & Tool Use

- GLM-4.5: —
- MiniMax-M2: 25.1 (#109)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| Terminal-Bench | — | 30% |
| Vending-Bench 2 | — | 160.6 |

## Reasoning

- GLM-4.5: 28.6 (#100)
- MiniMax-M2: 19.4 (#258)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| Kagi LLM Benchmark | 57.9% | 57.8% |
| LMArena Hard Prompts | 1429 | 1357 |
| NYT Connections (extended) | — | 14.8% |

## Math

- GLM-4.5: 39.0 (#116)
- MiniMax-M2: 37.3 (#160)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| LMArena Math | 1427 | 1352 |

## Knowledge

- GLM-4.5: 35.9 (#179)
- MiniMax-M2: 37.0 (#163)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| LMArena Expert | 1433 | 1337 |
| Humanity's Last Exam | 8.3% | — |
| Confabulations | 11.3% | — |

## Multilingual

- GLM-4.5: 52.8 (#77)
- MiniMax-M2: 45.3 (#171)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| LMArena Non-English | 1417 | 1313 |
| LMArena Chinese | 1465 | 1366 |
| LMArena French | 1418 | 1335 |
| LMArena German | 1407 | 1355 |
| LMArena Russian | 1414 | 1331 |
| LMArena Spanish | 1454 | 1326 |
| LMArena Japanese | 1415 | — |
| LMArena Korean | 1380 | — |

## Instruction Following

- GLM-4.5: 74.1 (#104)
- MiniMax-M2: 70.2 (#166)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| LMArena Instruction Following | 1404 | 1328 |

## Long Context

- GLM-4.5: 38.2 (#201)
- MiniMax-M2: 40.5 (#153)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| LMArena Longer Query | 1412 | 1331 |
| Fiction.LiveBench | 58.3% | — |

## Writing & Preference

- GLM-4.5: 57.5 (#127)
- MiniMax-M2: 53.0 (#162)

| Benchmark | GLM-4.5 | MiniMax-M2 |
|---|---|---|
| LMArena Text | 1430 | 1340 |
| LMArena Creative Writing | 1395 | 1286 |
| LMArena Multi-Turn | 1415 | 1361 |
| Short-Story Creative Writing | 73.4% | — |
| EQ-Bench Creative Writing | 1343 | — |

## FAQ

### Is GLM-4.5 better than MiniMax-M2?

GLM-4.5 is the stronger model overall, scoring 42.0 to 37.4 on the Noometry Index. MiniMax-M2 costs 1.9× less per token, which makes it the better buy when GLM-4.5's lead doesn't matter for your workload.

### Which is cheaper, GLM-4.5 or MiniMax-M2?

MiniMax-M2 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; GLM-4.5 lists at $0.60 and $2.20.

### Is GLM-4.5 or MiniMax-M2 better for coding?

GLM-4.5 scores higher on coding benchmarks: 41.4 versus 39.3 in the Noometry coding category.

### Which has the bigger context window?

MiniMax-M2 does, with 205K tokens against 131K.

### How many benchmarks do GLM-4.5 and MiniMax-M2 share?

17 benchmarks have published results for both models. GLM-4.5 has 27 scored results on Noometry and MiniMax-M2 has 21.