Modelos LLM
20 modelos
Datos: OpenRouter y fuentes oficiales · actualización diaria · al 25/9/2026
| # | Modelo ↕ | Proveedor ↕ | Puntuación ↕? | Contexto ↕? | Velocidad ↕? | MMLU ↕? | GPQA ↕? | HumanEval ↕? | SWE-B ↕? | Entrada $/1M ↕? | Salida $/1M ↕? | ⭐ Top ↑? | OSS ↕? | Lanzado ↕? |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | DeepSeek | - | 1M | - | - | - | - | - | $0.220 | $0.660 | - | - | ago 2026 | |
2 | DeepSeek | - | 1M | - | - | - | - | - | $0.462 | $1.39 | - | - | ago 2026 | |
3 | DeepSeek | - | 1M | - | - | - | - | - | $0.030 | $0.320 | - | - | ago 2026 | |
4 | DeepSeek | - | 1M | - | - | - | - | - | $0.030 | $0.320 | - | - | jul 2026 | |
5 | DeepSeek | - | 1M | - | - | - | - | - | $0.816 | $1.63 | - | - | abr 2026 | |
6 | DeepSeek | - | 1M | - | - | - | - | - | $0.049 | $0.098 | - | - | abr 2026 | |
7 | DeepSeek | - | 131K | - | - | - | - | - | $0.269 | $0.400 | - | - | dic 2025 | |
8 | DeepSeek | - | 164K | - | - | - | - | - | $0.270 | $0.410 | - | - | sept 2025 | |
9 | DeepSeek | - | 164K | - | - | - | - | - | $0.270 | $1.00 | - | - | sept 2025 | |
10 | DeepSeek | - | 164K | - | - | - | - | - | $0.250 | $0.950 | - | - | ago 2025 | |
11 | DeepSeek | - | 164K | - | - | - | - | - | $0.500 | $2.15 | - | - | may 2025 | |
12 | DeepSeek | - | 164K | - | - | - | - | - | $0.250 | $1.00 | - | - | mar 2025 | |
13 | DeepSeek | - | 128K | - | - | - | - | - | $0.290 | $0.290 | - | ✓ OSS | ene 2025 | |
14 | DeepSeek | - | 131K | - | - | - | - | - | $0.800 | $0.800 | - | ✓ OSS | ene 2025 | |
15 | DeepSeek | 75 | 64K | 30 | 90.8 | 71.5 | 92.6 | 49.2 | $0.700 | $2.50 | - | ✓ OSS | ene 2025 | |
16 | DeepSeek | 68 | 64K | 45 | 88.5 | 59.1 | 89.4 | 42.0 | $0.320 | $0.890 | - | ✓ OSS | - | |
17 | DeepSeek | - | 1M | - | - | - | - | - | $0.112 | $0.336 | - | - | - | |
18 | DeepSeek | - | 1M | - | - | - | - | - | $0.040 | $0.490 | - | - | - | |
19 | DeepSeek | - | 164K | - | - | - | - | - | $0.287 | $0.431 | - | - | - | |
20 | DeepSeek | - | 1M | - | - | - | - | - | $0.350 | $1.05 | - | - | - |
Puntuación = MMLU×20% + GPQA×30% + HumanEval×25% + SWE-Bench×25%⭐ Top - valoración media de usuarios TECHAGENTtok/s - tokens de salida/seg vía API
¿No encuentras un modelo que conoces? Inicia sesión para sugerirlo.