Up to four models
Qwen3.7-Max vs GLM-5.2
On the index (GPT-4 = 100), Qwen3.7-Max scores 160 and GLM-5.2 154. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
160144–177
WeightsClosed
ReleasedMay 2026
Price in / out$1.48 / $4.42
Context1M
Core tests taken4 of 9
154138–169
WeightsOpen
ReleasedJun 2026
Price in / out$0.0070 / $6.8
Context1.05M
Core tests taken6 of 9
Qwen3.7-Max leads GLM-5.2 by 7 points, ahead on 1 of 4 shared core tests, at a similar price.
Test by testthird-party results
Every result as a table
| Test | Qwen3.7-Max | GLM-5.2 |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 90.9% | 91.9% |
| LiveBench Data Analysis | 71.8% | 73.7% |
| LiveBench IF | 74.0% | 62.3% |
| LiveBench Language | 79.7% | 76.2% |
| LiveBench Reasoning | 83.3% | 78.6% |
| GPQA Diamond (Vals AI) | 90.1% | 85.6% |
| MMLU Pro | 89.3% | 86.7% |
| EBR-bench | 9.5% | 9.5% |
| SimpleBench | 70.4% | 58.8% |
| SimpleQA Verified | 55.8% | 34.2% |
| Maths | ||
| FrontierMath (tiers 1-3) | 64.6% | 59.2% |
| AIME-style maths (OTIS mock) | 95.6% | 86.4% |
| LiveBench Mathematics | 85.3% | 89.8% |
| Chess puzzles | 27.0% | 22.0% |
| FrontierMath (tier 4) | 34.2% | 29.3% |
| Coding | ||
| SWE-bench Verified | 77.3% | 78.7% |
| LiveBench Agentic Coding | 43.6% | 51.8% |
| LiveBench Coding | 74.2% | 79.7% |
| Code Migration | 13.1% | 37.9% |
| LiveCodeBench | 87.1% | 69.5% |
| SWE-bench | 68.8% | 82.8% |
| Vibe Code Bench v1.1 | 47.7% | 64.0% |
| Agents | ||
| DTBench | 92.3% | 93.6% |
| LMCA | 44.0% | 45.8% |
| Mystery game puzzles | 32.0% | 19.0% |
| Professional work | ||
| Finance Agent (v2) | 47.8% | 49.7% |
| CorpFin v2 | 63.7% | 66.1% |
| EMB | 57.0% | 61.5% |
| LegalBench | 84.9% | 84.1% |
| Legal Research Bench | 25.5% | 31.3% |
| MedCode | 38.8% | 40.8% |
| MedScribe | 79.4% | 83.5% |
| TaxEval v2 | 75.3% | 73.3% |
Tests at least two of these models have taken. Bold rows are the core tests.