DeepSeek V3.2 vs MiMo-V2-Flash
Lado a lado. Cada métrica. Cada benchmark.
| Tipo | DeepSeek V3.2 | MiMo-V2-Flash |
|---|---|---|
| Provider |  DeepSeek |  xiaomi |
| pontuação média | 53.0 | 73.3 |
| Preço de entrada | $0.26 | $0.09 |
| Preço de saída | $0.38 | $0.29 |
| Janela de contexto | 164K tokens (~82 books) | 262K tokens (~131 books) |
| Lançado em | 2025-12-01 | 2025-12-14 |
| Código aberto | Open Source | Open Source |
Pontuações de benchmark
11 benchmarks · DeepSeek V3.2: 10, MiMo-V2-Flash: 1
| Benchmark | Categoria | DeepSeek V3.2 | MiMo-V2-Flash |
|---|---|---|---|
| Artificial Analysis — Agentic Index | speed | 52.9 | 48.8 |
| Artificial Analysis — Coding Index | speed | 36.7 | 33.5 |
| Artificial Analysis — Quality Index | speed | 41.7 | 41.5 |
| Chatbot Arena Elo — Coding | arena | 1326.9 | 1336.5 |
| Chatbot Arena Elo — Overall | arena | 1424.4 | 1392.1 |
| OpenCompass — AIME2025 | math | 93.0 | 92.9 |
| OpenCompass — GPQA-Diamond | knowledge | 84.6 | 82.1 |
| OpenCompass — HLE | knowledge | 23.2 | 20.5 |
| OpenCompass — IFEval | language | 89.7 | 89.5 |
| OpenCompass — LiveCodeBenchV6 | coding | 75.4 | 71.8 |
| OpenCompass — MMLU-Pro | knowledge | 85.8 | 83.1 |