Model comparison
GLM-4.5-Air vs Qwen2.5 72B Instruct
GLM-4.5-Air is the stronger model overall, scoring 38.9 to 31.9 on the Noometry Index.
Last verified . 23 shared benchmarks.
Alibaba (Qwen)Summary
- They share 23 benchmarks with published results for both. GLM-4.5-Air scores higher in 8 categories and Qwen2.5 72B Instruct in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-4.5-Air leads 36.2 to 19.3.
- The biggest single-benchmark swing is GPQA (HELM): 59.4% for GLM-4.5-Air and 42.6% for Qwen2.5 72B Instruct.
- GLM-4.5-Air is cheaper at $0.20 / $1.10 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
Side by side
| GLM-4.5-Air | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Alibaba (Qwen) |
| Noometry Index | 38.9 | 31.9 |
| Released | 2025-07-20 | 2024-09 |
| Weights | Open | Open |
| Context window | 131K | 131K |
| Max output | 98K | 8K |
| Input $ / M tokens | $0.20 | $1.40 |
| Output $ / M tokens | $1.10 | $5.60 |
| Results tracked | 27 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GLM-4.5-Air: 33.3 (#259), Qwen2.5 72B Instruct: 33.2 (#260)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Coding | 1397 | 1292 |
| GSO | 2.9% | — |
| WeirdML | — | 16% |
| BigCodeBench Instruct | — | 45.8% |
| BigCodeBench Complete | — | 55.9% |
Agentic & Tool Use Not comparable
GLM-4.5-Air: —, Qwen2.5 72B Instruct: 22.1 (#133)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| TheAgentCompany | — | 5.7% |
| BALROG | — | 16.2% |
| METR Time Horizons | — | 35.8% |
Reasoning GLM-4.5-Air leads
GLM-4.5-Air: 24.1 (#166), Qwen2.5 72B Instruct: 22.3 (#199)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1379 | 1271 |
| ForecastBench | 59.2 | 57.5 |
| Kagi LLM Benchmark | 43% | — |
| DTBench | — | 62.9% |
| LMCA | — | 13.4% |
| BIG-Bench Hard | — | 79.8% |
| Epoch Capabilities Index | — | 129 |
| HellaSwag | — | 84.8% |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math GLM-4.5-Air leads
GLM-4.5-Air: 36.2 (#170), Qwen2.5 72B Instruct: 19.3 (#287)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| Omni-MATH | 39.1% | 33% |
| LMArena Math | 1396 | 1283 |
| OTIS Mock AIME 2024-2025 | — | 8.1% |
| MATH Level 5 | — | 63.2% |
Knowledge GLM-4.5-Air leads
GLM-4.5-Air: 35.0 (#191), Qwen2.5 72B Instruct: 27.0 (#253)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| MMLU-Pro | 76.2% | 63.1% |
| GPQA (HELM) | 59.4% | 42.6% |
| LMArena Expert | 1370 | 1245 |
| GPQA Diamond | — | 49.1% |
| Humanity's Last Exam | 8.1% | — |
| Confabulations | — | 19.1% |
| Vectara Hallucination Rate | 9.3% | — |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multilingual GLM-4.5-Air leads
GLM-4.5-Air: 49.1 (#135), Qwen2.5 72B Instruct: 41.0 (#213)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1366 | 1252 |
| LMArena Chinese | 1426 | 1272 |
| LMArena French | 1399 | 1280 |
| LMArena German | 1377 | 1234 |
| LMArena Japanese | 1348 | 1180 |
| LMArena Korean | 1308 | 1188 |
| LMArena Russian | 1373 | 1264 |
| LMArena Spanish | 1386 | 1256 |
Instruction Following GLM-4.5-Air leads
GLM-4.5-Air: 69.6 (#171), Qwen2.5 72B Instruct: 65.5 (#221)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| IFEval | 81.2% | 80.6% |
| LMArena Instruction Following | 1354 | 1254 |
Long Context GLM-4.5-Air leads
GLM-4.5-Air: 41.6 (#135), Qwen2.5 72B Instruct: 38.9 (#188)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1366 | 1282 |
Writing & Preference GLM-4.5-Air leads
GLM-4.5-Air: 55.9 (#139), Qwen2.5 72B Instruct: 46.7 (#215)
| Benchmark | GLM-4.5-Air | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1384 | 1269 |
| LMArena Creative Writing | 1343 | 1221 |
| WildBench | 78.9% | 80.2% |
| LMArena Multi-Turn | 1371 | 1272 |
Frequently asked questions
Is GLM-4.5-Air better than Qwen2.5 72B Instruct?
GLM-4.5-Air is the stronger model overall, scoring 38.9 to 31.9 on the Noometry Index.
Which is cheaper, GLM-4.5-Air or Qwen2.5 72B Instruct?
GLM-4.5-Air is cheaper. It lists at $0.20 per million input tokens and $1.10 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is GLM-4.5-Air or Qwen2.5 72B Instruct better for coding?
They score almost the same on coding (33.3 vs 33.2); test both on your own repository before choosing.
Which has the bigger context window?
Both accept 131K tokens.
How many benchmarks do GLM-4.5-Air and Qwen2.5 72B Instruct share?
23 benchmarks have published results for both models. GLM-4.5-Air has 27 scored results on Noometry and Qwen2.5 72B Instruct has 43.