Model comparison
GLM-5.2 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 51.1 on the Noometry Index.
Last verified . 35 shared benchmarks.
Alibaba (Qwen)Summary
- They share 35 benchmarks with published results for both. GLM-5.2 scores higher in 1 category and Qwen3.8 Max in 8 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 55.7.
- The biggest single-benchmark swing is ProofBench: 35% for GLM-5.2 and 58% for Qwen3.8 Max.
- GLM-5.2 is cheaper at $1.40 / $4.40 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- GLM-5.2 has downloadable open weights; the other is API-only.
Side by side
| GLM-5.2 | Qwen3.8 Max | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Alibaba (Qwen) |
| Noometry Index | 51.1 | 56.8 |
| Released | 2026-06-13 | 2026-08-02 |
| Weights | Open | Proprietary |
| Context window | 1M | 1M |
| Max output | 131K | 131K |
| Input $ / M tokens | $1.40 | $2 |
| Output $ / M tokens | $4.40 | $6 |
| Results tracked | 51 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
GLM-5.2: 51.3 (#41), Qwen3.8 Max: 53.5 (#29)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| DeepSWE | 43.8% | 57.5% |
| LMArena WebDev | 1603 | 1674 |
| SciCode | 50.5% | 53.2% |
| LMArena Coding | 1485 | 1502 |
| SWE-bench Verified | 78.7% | — |
| FrontierCode | 24.5% | — |
| FrontierSWE | — | 17.8% |
| WeirdML | 70.1% | — |
| ALE-Bench | 1,047 | — |
Agentic & Tool Use Qwen3.8 Max leads
GLM-5.2: 32.4 (#63), Qwen3.8 Max: 45.4 (#14)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | 45.2% | 63.3% |
| τ²-bench Banking | 37.1% | 55.1% |
| PostTrainBench | 31.7% | — |
| GBAEval | 0% | — |
| GDP.pdf | — | 23.2% |
| Vending-Bench 2 | 8,314 | — |
Reasoning Qwen3.8 Max leads
GLM-5.2: 42.3 (#52), Qwen3.8 Max: 54.4 (#26)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| NYT Connections (extended) | 74.3% | 88.3% |
| CritPt | 20.9% | 20% |
| Chess Puzzles | 21% | 40% |
| LMArena Hard Prompts | 1480 | 1496 |
| Mystery Game Puzzles | 19% | 38% |
| DTBench | 93.6% | 92% |
| LMCA | 45.8% | 46.2% |
| Epoch Capabilities Index | 151.78 | 156.41 |
| ARC-AGI-2 | 22.8% | — |
| SimpleBench | 58.8% | — |
| Kagi LLM Benchmark | 62.6% | — |
| ARC-AGI-1 | 77% | — |
| EBR-Bench | 9.5% | — |
| Surface Evolver Bench | 55.6% | — |
Math Qwen3.8 Max leads
GLM-5.2: 55.7 (#43), Qwen3.8 Max: 73.2 (#20)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.2% | 74.7% |
| FrontierMath Tier 4 | 29.3% | 46.3% |
| OTIS Mock AIME 2024-2025 | 86.4% | 100% |
| ProofBench | 35% | 58% |
| LMArena Math | 1482 | 1499 |
| MathArena Final-Answer Competitions | 67.6% | — |
Knowledge Qwen3.8 Max leads
GLM-5.2: 57.1 (#40), Qwen3.8 Max: 61.7 (#27)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 91.9% | 92.7% |
| SimpleQA Verified | 34.2% | 47.3% |
| LMArena Expert | 1486 | 1507 |
Multimodal Not comparable
GLM-5.2: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Too close to call
GLM-5.2: 55.8 (#26), Qwen3.8 Max: 56.7 (#18)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1459 | 1472 |
| LMArena Chinese | 1519 | 1538 |
| LMArena French | 1479 | 1503 |
| LMArena German | 1468 | 1483 |
| LMArena Japanese | 1451 | 1467 |
| LMArena Korean | 1445 | 1461 |
| LMArena Russian | 1466 | 1481 |
| LMArena Spanish | 1477 | 1492 |
Instruction Following Too close to call
GLM-5.2: 76.9 (#34), Qwen3.8 Max: 77.6 (#17)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1465 | 1479 |
Long Context Too close to call
GLM-5.2: 45.3 (#43), Qwen3.8 Max: 45.6 (#31)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1479 | 1489 |
Writing & Preference GLM-5.2 leads
GLM-5.2: 70.4 (#21), Qwen3.8 Max: 67.1 (#30)
| Benchmark | GLM-5.2 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1470 | 1483 |
| LMArena Creative Writing | 1462 | 1479 |
| LMArena Multi-Turn | 1469 | 1489 |
| EQ-Bench Creative Writing | 1757 | — |
| EQ-Bench 4 | 1222 | — |
Frequently asked questions
Is GLM-5.2 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 51.1 on the Noometry Index.
Which is cheaper, GLM-5.2 or Qwen3.8 Max?
GLM-5.2 is cheaper. It lists at $1.40 per million input tokens and $4.40 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is GLM-5.2 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 51.3 in the Noometry coding category.
Which has the bigger context window?
Both accept 1M tokens.
How many benchmarks do GLM-5.2 and Qwen3.8 Max share?
35 benchmarks have published results for both models. GLM-5.2 has 51 scored results on Noometry and Qwen3.8 Max has 39.