DeepSeek V4.1 Flash vs. Hermes 4 - Llama-3.1 405B: comparación de lanzamientos
Comparación de los lanzamientos DeepSeek V4.1 Flash y Hermes 4 - Llama-3.1 405B: 2 y 2 modelos respectivamente, cada uno con distintas características de inteligencia, rendimiento y precio. A continuación se comparan las métricas clave de los 4 modelos.
- En inteligencia, DeepSeek V4.1 Flash obtiene la puntuación más alta: DeepSeek V4.1 Flash (Reasoning, Max Effort), con 39, frente a Hermes 4 - Llama-3.1 405B (Reasoning), con 7 de Hermes 4 - Llama-3.1 405B.
- En velocidad de salida, DeepSeek V4.1 Flash es el más rápido: DeepSeek V4.1 Flash (Non-Reasoning), con 296 tokens/s, frente a Hermes 4 - Llama-3.1 405B (Non-reasoning), con 42 tokens/s de Hermes 4 - Llama-3.1 405B.
Inteligencia
Artificial Analysis Intelligence Index
Intelligence Index vs. Cost per Intelligence Index Task
Cost per Task (USD, Log Scale)
Costo
Cost per Intelligence Index Task
Velocidad y latencia
Output Speed
Puntuaciones de capacidades
Índices de capacidades
Incorporates 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better
Incorporates 6 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, AutomationBench-AA, AA-LCR v1.1, GDP.pdf · Higher is better
Incorporates 7 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, Humanity's Last Exam, AA-LCR v1.1, GDP.pdf, AutomationBench-AA · Higher is better
Incorporates 6 evaluations: AA-Omniscience, GDPval-AA v2.1, AA-Briefcase v1.1, MLCR-AA, Humanity's Last Exam, AutomationBench-AA · Higher is better
Incorporates 6 evaluations: AA-Omniscience, Humanity's Last Exam, CritPt, GDPval-AA v2.1, AA-Briefcase v1.1, Terminal-Bench 4.0 · Higher is better
Incorporates 5 evaluations: AA-Omniscience, Humanity's Last Exam, GDPval-AA v2.1, AA-Briefcase v1.1, AA-LCR v1.1 · Higher is better
Más detalles
Modalidad de entrada | Modalidad de salida | Pesos | Benchmarks de proveedores | ||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4.1 Flash (Reasoning, Max Effort) |  DeepSeek | 39 | 45 | 52 | 41 | 41 | 39 | 44 | $0.27 | 0,2 US$ | $0.30 | $1.20 | $0.006 | 477 US$ | 89k | 63k | 253M | 222 | 0.92s | 9.94s | 12.20s | 304.19s | 552B 16B activos en inferencia | 1M | sept 2026 | - | Sí | Admite: texto e imágenes | Admite: texto |    +15 | |||
| DeepSeek V4.1 Flash (Non-Reasoning) | DeepSeek | 25 | 30 | 36 | 30 | - | 25 | 32 | $0.15 | 0,2 US$ | $0.30 | $1.20 | $0.006 | 156 US$ | 34k | 0 | 45M | 296 | 1.16s | 1.16s | 2.85s | 116.47s | 552B 16B activos en inferencia | 1M | sept 2026 | - | No | Admite: texto e imágenes | Admite: texto | | |||
| Hermes 4 - Llama-3.1 405B (Reasoning) |  Nous Research | 7 | - | - | - | - | - | - | - | 1,2 US$ | $1.00 | $3.00 | - | - | - | - | - | 42 | 2.37s | 50.55s | 62.59s | - | 406B | 128k | ago 2025 | - | Sí | Admite: texto | Admite: texto |  | |||
| Hermes 4 - Llama-3.1 405B (Non-reasoning) | Nous Research | 7 | - | - | - | - | - | - | - | 1,2 US$ | $1.00 | $3.00 | - | - | - | - | - | 42 | 2.40s | 2.40s | 14.43s | - | 406B | 128k | ago 2025 | - | No | Admite: texto | Admite: texto | | |||