# GLM-4.5 vs GPT-4

> GLM-4.5 is the stronger model overall, scoring 42.0 to 29.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/glm-4-5-vs-gpt-4
- Last updated: 2026-10-11
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. GLM-4.5 scores higher in 8 categories and GPT-4 in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-4.5 leads 39.0 to 10.8.
- The biggest single-benchmark swing is WeirdML: 40.6% for GLM-4.5 and 12.4% for GPT-4.
- GLM-4.5 is cheaper at $0.60 / $2.20 per million input/output tokens, against $30 / $60 for GPT-4.
- GLM-4.5 accepts more context: 131K tokens versus 8K.
- GLM-4.5 has downloadable open weights; the other is API-only.

## Snapshot

| | GLM-4.5 | GPT-4 |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 42.0 | 29.1 |
| Rank | 122 | 316 |
| Context | 131K | 8K |
| Input $/M | $0.60 | $30 |
| Output $/M | $2.20 | $60 |
| Weights | Open | Proprietary |

## Coding

- GLM-4.5: 41.4 (#125)
- GPT-4: 31.6 (#283)

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| WeirdML | 40.6% | 12.4% |
| LMArena Coding | 1434 | 1254 |
| SWE-bench Verified (bash only) | 54.2% | — |
| BigCodeBench Instruct | — | 46% |
| BigCodeBench Complete | — | 57.2% |
| ALE-Bench | 344.82 | — |
| AlgoTune | 1.52 | — |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- GLM-4.5: —
- GPT-4: —

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| METR Time Horizons | — | 36.1% |

## Reasoning

- GLM-4.5: 28.6 (#100)
- GPT-4: 17.8 (#289)

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| LMArena Hard Prompts | 1429 | 1241 |
| Kagi LLM Benchmark | 57.9% | — |
| Chess Puzzles | — | 4% |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 62.7% |
| LMCA | — | 17.1% |
| BIG-Bench Hard | — | 75.1% |
| Epoch Capabilities Index | — | 125.89 |
| ForecastBench | — | 57.8 |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |

## Math

- GLM-4.5: 39.0 (#116)
- GPT-4: 10.8 (#309)

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| LMArena Math | 1427 | 1269 |
| OTIS Mock AIME 2024-2025 | — | 1.1% |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |

## Knowledge

- GLM-4.5: 35.9 (#179)
- GPT-4: 18.4 (#282)

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| LMArena Expert | 1433 | 1211 |
| GPQA Diamond | — | 35.7% |
| Humanity's Last Exam | 8.3% | — |
| Confabulations | 11.3% | — |
| MMLU | — | 86.4% |
| TriviaQA | — | 84.8% |

## Multilingual

- GLM-4.5: 52.8 (#77)
- GPT-4: 40.6 (#215)

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| LMArena Non-English | 1417 | 1246 |
| LMArena Chinese | 1465 | 1242 |
| LMArena French | 1418 | 1283 |
| LMArena German | 1407 | 1251 |
| LMArena Japanese | 1415 | 1209 |
| LMArena Korean | 1380 | 1184 |
| LMArena Russian | 1414 | 1251 |
| LMArena Spanish | 1454 | 1261 |

## Instruction Following

- GLM-4.5: 74.1 (#104)
- GPT-4: 65.3 (#222)

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1404 | 1241 |

## Long Context

- GLM-4.5: 38.2 (#201)
- GPT-4: 37.7 (#212)

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1412 | 1244 |
| Fiction.LiveBench | 58.3% | — |

## Writing & Preference

- GLM-4.5: 57.5 (#127)
- GPT-4: 34.9 (#268)

| Benchmark | GLM-4.5 | GPT-4 |
|---|---|---|
| LMArena Text | 1430 | 1263 |
| LMArena Creative Writing | 1395 | 1244 |
| EQ-Bench Creative Writing | 1343 | 752 |
| LMArena Multi-Turn | 1415 | 1257 |
| Short-Story Creative Writing | 73.4% | — |

## FAQ

### Is GLM-4.5 better than GPT-4?

GLM-4.5 is the stronger model overall, scoring 42.0 to 29.1 on the Noometry Index.

### Which is cheaper, GLM-4.5 or GPT-4?

GLM-4.5 is cheaper. It lists at $0.60 per million input tokens and $2.20 per million output tokens; GPT-4 lists at $30 and $60.

### Is GLM-4.5 or GPT-4 better for coding?

GLM-4.5 scores higher on coding benchmarks: 41.4 versus 31.6 in the Noometry coding category.

### Which has the bigger context window?

GLM-4.5 does, with 131K tokens against 8K.

### How many benchmarks do GLM-4.5 and GPT-4 share?

19 benchmarks have published results for both models. GLM-4.5 has 27 scored results on Noometry and GPT-4 has 38.