Model comparison
GLM-5.3 vs Qwen1.5-72B
GLM-5.3 is the stronger model overall, scoring 54.8 to 30.8 on the Noometry Index.
Last verified . 18 shared benchmarks.
Alibaba (Qwen)Summary
- They share 18 benchmarks with published results for both. GLM-5.3 scores higher in 8 categories and Qwen1.5-72B in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5.3 leads 58.3 to 11.5.
- The biggest single-benchmark swing is GPQA Diamond: 90.9% for GLM-5.3 and 28.8% for Qwen1.5-72B.
Side by side
| GLM-5.3 | Qwen1.5-72B | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Alibaba (Qwen) |
| Noometry Index | 54.8 | 30.8 |
| Released | 2026-08-14 | 2024-02-04 |
| Weights | Open | Open |
| Context window | 1M | — |
| Max output | 131K | — |
| Input $ / M tokens | $1.40 | — |
| Output $ / M tokens | $4.40 | — |
| Results tracked | 42 | 22 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3 leads
GLM-5.3: 59.5 (#14), Qwen1.5-72B: 31.9 (#277)
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| LMArena Coding | 1496 | 1165 |
| DeepSWE | 69% | — |
| FrontierCode | 40.1% | — |
| CursorBench | 42.6% | — |
| LMArena WebDev | 1622 | — |
| FrontierSWE | 30.2% | — |
| SciCode | 59% | — |
| WeirdML | 75.4% | — |
| BigCodeBench Instruct | — | 33.2% |
| BigCodeBench Complete | — | 40.3% |
| ALE-Bench | 1,317 | — |
| HumanEval+ | — | 59.1% |
| MBPP+ | — | 61.6% |
Agentic & Tool Use Not comparable
GLM-5.3: 36.4 (#38), Qwen1.5-72B: —
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| APEX-Agents | 56.6% | — |
| Vending-Bench 2 | 8,164 | — |
Reasoning GLM-5.3 leads
GLM-5.3: 46.1 (#46), Qwen1.5-72B: 22.2 (#203)
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| LMArena Hard Prompts | 1489 | 1148 |
| NYT Connections (extended) | 74.2% | — |
| CritPt | 19.1% | — |
| Chess Puzzles | 21% | — |
| Mystery Game Puzzles | 33% | — |
| DTBench | 87.7% | — |
| LMCA | 55.5% | — |
| Bench to the Future 3 | 0.15 | — |
| Epoch Capabilities Index | 155.61 | — |
Math GLM-5.3 leads
GLM-5.3: 62.3 (#33), Qwen1.5-72B: 33.2 (#205)
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| LMArena Math | 1489 | 1164 |
| FrontierMath (Tiers 1-3) | 68.8% | — |
| FrontierMath Tier 4 | 29.3% | — |
| OTIS Mock AIME 2024-2025 | 91.1% | — |
| ProofBench | 49% | — |
Knowledge GLM-5.3 leads
GLM-5.3: 58.3 (#37), Qwen1.5-72B: 11.5 (#300)
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| GPQA Diamond | 90.9% | 28.8% |
| LMArena Expert | 1516 | 1136 |
| SimpleQA Verified | 41% | — |
Multilingual GLM-5.3 leads
GLM-5.3: 55.7 (#28), Qwen1.5-72B: 33.2 (#253)
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| LMArena Non-English | 1457 | 1135 |
| LMArena Chinese | 1528 | 1186 |
| LMArena French | 1499 | 1159 |
| LMArena German | 1499 | 1084 |
| LMArena Japanese | 1453 | 1061 |
| LMArena Korean | 1472 | 1050 |
| LMArena Russian | 1463 | 1104 |
| LMArena Spanish | 1460 | 1110 |
Instruction Following GLM-5.3 leads
GLM-5.3: 77.5 (#23), Qwen1.5-72B: 59.3 (#256)
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| LMArena Instruction Following | 1477 | 1141 |
Long Context GLM-5.3 leads
GLM-5.3: 45.4 (#41), Qwen1.5-72B: 35.1 (#243)
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| LMArena Longer Query | 1482 | 1157 |
Writing & Preference GLM-5.3 leads
GLM-5.3: 75.7 (#6), Qwen1.5-72B: 37.3 (#258)
| Benchmark | GLM-5.3 | Qwen1.5-72B |
|---|---|---|
| LMArena Text | 1471 | 1166 |
| LMArena Creative Writing | 1457 | 1137 |
| LMArena Multi-Turn | 1472 | 1160 |
| EQ-Bench Creative Writing | 2075 | — |
Frequently asked questions
Is GLM-5.3 better than Qwen1.5-72B?
GLM-5.3 is the stronger model overall, scoring 54.8 to 30.8 on the Noometry Index.
Is GLM-5.3 or Qwen1.5-72B better for coding?
GLM-5.3 scores higher on coding benchmarks: 59.5 versus 31.9 in the Noometry coding category.
How many benchmarks do GLM-5.3 and Qwen1.5-72B share?
18 benchmarks have published results for both models. GLM-5.3 has 42 scored results on Noometry and Qwen1.5-72B has 22.