Model comparison
GLM-5 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 46.1 on the Noometry Index. GLM-5 costs 1.9× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
Alibaba (Qwen)Summary
- They share 24 benchmarks with published results for both. GLM-5 scores higher in 0 categories and Qwen3.8 Max in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 46.4.
- The biggest single-benchmark swing is τ²-bench Banking: 9.8% for GLM-5 and 55.1% for Qwen3.8 Max.
- GLM-5 is cheaper at $1 / $3.20 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- Qwen3.8 Max accepts more context: 1M tokens versus 205K.
- GLM-5 has downloadable open weights; the other is API-only.
Side by side
| GLM-5 | Qwen3.8 Max | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Alibaba (Qwen) |
| Noometry Index | 46.1 | 56.8 |
| Released | 2026-02-11 | 2026-08-02 |
| Weights | Open | Proprietary |
| Context window | 205K | 1M |
| Max output | 131K | 131K |
| Input $ / M tokens | $1 | $2 |
| Output $ / M tokens | $3.20 | $6 |
| Results tracked | 45 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
GLM-5: 49.0 (#52), Qwen3.8 Max: 53.5 (#29)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| LMArena WebDev | 1434 | 1674 |
| LMArena Coding | 1461 | 1502 |
| SWE-bench Verified | 72.1% | — |
| DeepSWE | — | 57.5% |
| SWE-bench Verified (bash only) | 72.8% | — |
| SWE-bench Multilingual | 69.7% | — |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| WeirdML | 48.2% | — |
| ALE-Bench | 765.62 | — |
Agentic & Tool Use Qwen3.8 Max leads
GLM-5: 31.1 (#71), Qwen3.8 Max: 45.4 (#14)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| τ²-bench Banking | 9.8% | 55.1% |
| Terminal-Bench | 52.4% | — |
| APEX-Agents | — | 63.3% |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Retail | 73.7% | — |
| τ²-bench Telecom | 86.8% | — |
| GDP.pdf | — | 23.2% |
| Vending-Bench 2 | 4,432 | — |
Reasoning Qwen3.8 Max leads
GLM-5: 27.6 (#116), Qwen3.8 Max: 54.4 (#26)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| NYT Connections (extended) | 74.8% | 88.3% |
| Chess Puzzles | 10% | 40% |
| LMArena Hard Prompts | 1452 | 1496 |
| Epoch Capabilities Index | 145.83 | 156.41 |
| ARC-AGI-2 | 4.9% | — |
| SimpleBench | 53.2% | — |
| Kagi LLM Benchmark | 75% | — |
| ARC-AGI-1 | 44.7% | — |
| CritPt | — | 20% |
| Mystery Game Puzzles | — | 38% |
| DTBench | — | 92% |
| LMCA | — | 46.2% |
| ForecastBench | 61 | — |
Math Qwen3.8 Max leads
GLM-5: 46.4 (#71), Qwen3.8 Max: 73.2 (#20)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 80% | 100% |
| LMArena Math | 1440 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| MathArena Final-Answer Competitions | 65.7% | — |
| ProofBench | — | 58% |
| FrontierMath (Feb 2025 set) | 16.4% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Qwen3.8 Max leads
GLM-5: 52.3 (#64), Qwen3.8 Max: 61.7 (#27)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 87.8% | 92.7% |
| LMArena Expert | 1454 | 1507 |
| SimpleQA Verified | — | 47.3% |
| Vectara Hallucination Rate | 10.1% | — |
Multimodal Not comparable
GLM-5: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
GLM-5: 53.7 (#58), Qwen3.8 Max: 56.7 (#18)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1430 | 1472 |
| LMArena Chinese | 1511 | 1538 |
| LMArena French | 1455 | 1503 |
| LMArena German | 1445 | 1483 |
| LMArena Japanese | 1416 | 1467 |
| LMArena Korean | 1423 | 1461 |
| LMArena Russian | 1436 | 1481 |
| LMArena Spanish | 1454 | 1492 |
Instruction Following Qwen3.8 Max leads
GLM-5: 75.2 (#67), Qwen3.8 Max: 77.6 (#17)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1428 | 1479 |
Long Context Too close to call
GLM-5: 44.7 (#60), Qwen3.8 Max: 45.6 (#31)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1446 | 1489 |
| CL-bench | 18.7% | — |
Writing & Preference Qwen3.8 Max leads
GLM-5: 66.0 (#38), Qwen3.8 Max: 67.1 (#30)
| Benchmark | GLM-5 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1446 | 1483 |
| LMArena Creative Writing | 1439 | 1479 |
| LMArena Multi-Turn | 1456 | 1489 |
| EQ-Bench Creative Writing | 1601 | — |
Frequently asked questions
Is GLM-5 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 46.1 on the Noometry Index. GLM-5 costs 1.9× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Which is cheaper, GLM-5 or Qwen3.8 Max?
GLM-5 is cheaper. It lists at $1 per million input tokens and $3.20 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is GLM-5 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 49.0 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 Max does, with 1M tokens against 205K.
How many benchmarks do GLM-5 and Qwen3.8 Max share?
24 benchmarks have published results for both models. GLM-5 has 45 scored results on Noometry and Qwen3.8 Max has 39.