Kimi K2.5 vs GLM 4.7
Côte à côte. Chaque métrique. Chaque benchmark.
| Type | Kimi K2.5 | GLM 4.7 |
|---|---|---|
| Provider |  moonshotai |  z-ai |
| score moyen | 52.0 | 50.5 |
| Prix d'entrée | $0.38 | $0.39 |
| Prix de sortie | $1.72 | $1.75 |
| Fenêtre de contexte | 262K tokens (~131 books) | 203K tokens (~101 books) |
| Sorti le | 2026-01-27 | 2025-12-22 |
| Code source ouvert | Open Source | Open Source |
Scores de benchmark
16 benchmarks · Kimi K2.5: 13, GLM 4.7: 3
| Benchmark | Catégorie | Kimi K2.5 | GLM 4.7 |
|---|---|---|---|
| APEX-Agents | agentic | 14.4 | 3.1 |
| Chess Puzzles | knowledge | 12.0 | 6.0 |
| FrontierMath-2025-02-28-Private | math | 27.9 | 2.4 |
| FrontierMath-Tier-4-2025-07-01-Private | math | 4.2 | 0.1 |
| GPQA diamond | knowledge | 83.5 | 77.8 |
| OpenCompass — AIME2025 | math | 91.9 | 95.4 |
| OpenCompass — GPQA-Diamond | knowledge | 88.1 | 86.9 |
| OpenCompass — HLE | knowledge | 28.6 | 25.4 |
| OpenCompass — IFEval | language | 93.9 | 90.2 |
| OpenCompass — LiveCodeBenchV6 | coding | 80.6 | 83.8 |
| OpenCompass — MMLU-Pro | knowledge | 86.2 | 84.0 |
| OTIS Mock AIME 2024-2025 | math | 92.2 | 83.3 |
| PostTrainBench | knowledge | 10.3 | 7.5 |
| SimpleBench | reasoning | 36.2 | 37.2 |
| SimpleQA Verified | knowledge | 33.9 | 31.5 |
| Terminal Bench | coding | 43.2 | 33.4 |