Kimi K2.5 vs GLM 5
Côte à côte. Chaque métrique. Chaque benchmark.
| Type | Kimi K2.5 | GLM 5 |
|---|---|---|
| Provider |  moonshotai |  z-ai |
| score moyen | 52.0 | 57.6 |
| Prix d'entrée | $0.38 | $0.72 |
| Prix de sortie | $1.72 | $2.30 |
| Fenêtre de contexte | 262K tokens (~131 books) | 80K tokens (~40 books) |
| Sorti le | 2026-01-27 | 2026-02-11 |
| Code source ouvert | Open Source | Open Source |
Scores de benchmark
18 benchmarks · Kimi K2.5: 11, GLM 5: 7
| Benchmark | Catégorie | Kimi K2.5 | GLM 5 |
|---|---|---|---|
| ARC-AGI | reasoning | 65.3 | 44.7 |
| ARC-AGI-2 | reasoning | 11.8 | 4.9 |
| Chess Puzzles | knowledge | 12.0 | 10.0 |
| FrontierMath-2025-02-28-Private | math | 27.9 | 16.4 |
| FrontierMath-Tier-4-2025-07-01-Private | math | 4.2 | 2.1 |
| GPQA diamond | knowledge | 83.5 | 83.8 |
| OpenCompass — AIME2025 | math | 91.9 | 95.8 |
| OpenCompass — GPQA-Diamond | knowledge | 88.1 | 85.3 |
| OpenCompass — HLE | knowledge | 28.6 | 28.1 |
| OpenCompass — IFEval | language | 93.9 | 93.2 |
| OpenCompass — LiveCodeBenchV6 | coding | 80.6 | 86.2 |
| OpenCompass — MMLU-Pro | knowledge | 86.2 | 85.2 |
| OTIS Mock AIME 2024-2025 | math | 92.2 | 80.0 |
| PostTrainBench | knowledge | 10.3 | 13.9 |
| SimpleBench | reasoning | 36.2 | 43.8 |
| SWE-Bench verified | coding | 73.8 | 72.1 |
| Terminal Bench | coding | 43.2 | 52.4 |
| WeirdML | coding | 45.6 | 48.2 |