# GLM-4.5-Air vs GLM-5.2

> GLM-5.2 is the stronger model overall, scoring 50.9 to 38.7 on the Noometry Index. GLM-4.5-Air costs 5.1× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/glm-4-5-air-vs-glm-5-2
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. GLM-4.5-Air scores higher in 0 categories and GLM-5.2 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5.2 leads 57.1 to 34.6.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 43% for GLM-4.5-Air and 62.6% for GLM-5.2.
- GLM-4.5-Air is cheaper at $0.20 / $1.10 per million input/output tokens, against $1.40 / $4.40 for GLM-5.2.
- GLM-5.2 accepts more context: 1M tokens versus 131K.

## Snapshot

| | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| Provider | Z.ai (Zhipu) | Z.ai (Zhipu) |
| Noometry Index | 38.7 | 50.9 |
| Rank | 169 | 38 |
| Context | 131K | 1M |
| Input $/M | $0.20 | $1.40 |
| Output $/M | $1.10 | $4.40 |
| Weights | Open | Open |

## Coding

- GLM-4.5-Air: 34.5 (#225)
- GLM-5.2: 50.9 (#37)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| LMArena Coding | 1397 | 1484 |
| SWE-bench Verified | — | 78.7% |
| DeepSWE | — | 43.8% |
| FrontierCode | — | 24.5% |
| LMArena WebDev | — | 1604 |
| SciCode | — | 50.5% |
| GSO | 2.9% | — |
| WeirdML | — | 70.1% |
| ALE-Bench | — | 1,047 |

## Agentic & Tool Use

- GLM-4.5-Air: —
- GLM-5.2: 31.6 (#63)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 31.7% |
| GBAEval | — | 0% |
| Vending-Bench 2 | — | 8,314 |

## Reasoning

- GLM-4.5-Air: 23.3 (#166)
- GLM-5.2: 42.1 (#46)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| Kagi LLM Benchmark | 43% | 62.6% |
| LMArena Hard Prompts | 1379 | 1480 |
| ARC-AGI-2 | — | 22.8% |
| SimpleBench | — | 58.8% |
| NYT Connections (extended) | — | 74.3% |
| ARC-AGI-1 | — | 77% |
| CritPt | — | 20.9% |
| Chess Puzzles | — | 21% |
| EBR-Bench | — | 9.5% |
| Mystery Game Puzzles | — | 19% |
| DTBench | — | 93.6% |
| LMCA | — | 45.8% |
| Surface Evolver Bench | — | 55.6% |
| Epoch Capabilities Index | — | 151.76 |
| ForecastBench | 59.2 | — |

## Math

- GLM-4.5-Air: 33.8 (#166)
- GLM-5.2: 55.4 (#35)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| LMArena Math | 1396 | 1482 |
| FrontierMath (Tiers 1-3) | — | 59.2% |
| FrontierMath Tier 4 | — | 29.3% |
| MathArena Final-Answer Competitions | — | 67.6% |
| OTIS Mock AIME 2024-2025 | — | 86.4% |
| ProofBench | — | 35% |
| Omni-MATH | 39.1% | — |

## Knowledge

- GLM-4.5-Air: 34.6 (#184)
- GLM-5.2: 57.1 (#36)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| LMArena Expert | 1370 | 1484 |
| GPQA Diamond | — | 91.9% |
| Humanity's Last Exam | 8.1% | — |
| SimpleQA Verified | — | 34.2% |
| MMLU-Pro | 76.2% | — |
| Vectara Hallucination Rate | 9.3% | — |
| GPQA (HELM) | 59.4% | — |

## Multilingual

- GLM-4.5-Air: 49.4 (#130)
- GLM-5.2: 56.0 (#25)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| LMArena Non-English | 1366 | 1459 |
| LMArena Chinese | 1427 | 1518 |
| LMArena French | 1399 | 1476 |
| LMArena German | 1376 | 1470 |
| LMArena Japanese | 1348 | 1451 |
| LMArena Korean | 1306 | 1445 |
| LMArena Russian | 1373 | 1467 |
| LMArena Spanish | 1385 | 1477 |

## Instruction Following

- GLM-4.5-Air: 69.8 (#165)
- GLM-5.2: 77.1 (#32)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| LMArena Instruction Following | 1353 | 1465 |
| IFEval | 81.2% | — |

## Long Context

- GLM-4.5-Air: 41.6 (#130)
- GLM-5.2: 45.3 (#43)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| LMArena Longer Query | 1367 | 1479 |

## Writing & Preference

- GLM-4.5-Air: 56.1 (#134)
- GLM-5.2: 70.7 (#20)

| Benchmark | GLM-4.5-Air | GLM-5.2 |
|---|---|---|
| LMArena Text | 1384 | 1470 |
| LMArena Creative Writing | 1343 | 1465 |
| LMArena Multi-Turn | 1371 | 1468 |
| EQ-Bench Creative Writing | — | 1757 |
| WildBench | 78.9% | — |
| EQ-Bench 4 | — | 1222 |

## FAQ

### Is GLM-4.5-Air better than GLM-5.2?

GLM-5.2 is the stronger model overall, scoring 50.9 to 38.7 on the Noometry Index. GLM-4.5-Air costs 5.1× less per token, which makes it the better buy when GLM-5.2's lead doesn't matter for your workload.

### Which is cheaper, GLM-4.5-Air or GLM-5.2?

GLM-4.5-Air is cheaper. It lists at $0.20 per million input tokens and $1.10 per million output tokens; GLM-5.2 lists at $1.40 and $4.40.

### Is GLM-4.5-Air or GLM-5.2 better for coding?

GLM-5.2 scores higher on coding benchmarks: 50.9 versus 34.5 in the Noometry coding category.

### Which has the bigger context window?

GLM-5.2 does, with 1M tokens against 131K.

### How many benchmarks do GLM-4.5-Air and GLM-5.2 share?

18 benchmarks have published results for both models. GLM-4.5-Air has 27 scored results on Noometry and GLM-5.2 has 50.