GLM 4.7 vs Qwen3.5 397B A17B
Côte à côte. Chaque métrique. Chaque benchmark.
| Type | GLM 4.7 | Qwen3.5 397B A17B |
|---|---|---|
| Provider |  z-ai |  Alibaba Qwen |
| score moyen | 50.5 | 78.4 |
| Prix d'entrée | $0.39 | $0.39 |
| Prix de sortie | $1.75 | $2.34 |
| Fenêtre de contexte | 203K tokens (~101 books) | 262K tokens (~131 books) |
| Sorti le | 2025-12-22 | 2026-02-16 |
| Code source ouvert | Open Source | Open Source |
Scores de benchmark
8 benchmarks · GLM 4.7: 3, Qwen3.5 397B A17B: 5
| Benchmark | Catégorie | GLM 4.7 | Qwen3.5 397B A17B |
|---|---|---|---|
| Chatbot Arena Elo — Coding | arena | 1439.2 | 1386.1 |
| Chatbot Arena Elo — Overall | arena | 1442.7 | 1447.7 |
| OpenCompass — AIME2025 | math | 95.4 | 92.3 |
| OpenCompass — GPQA-Diamond | knowledge | 86.9 | 88.4 |
| OpenCompass — HLE | knowledge | 25.4 | 27.5 |
| OpenCompass — IFEval | language | 90.2 | 91.5 |
| OpenCompass — LiveCodeBenchV6 | coding | 83.8 | 83.0 |
| OpenCompass — MMLU-Pro | knowledge | 84.0 | 87.6 |