Análisis de velocidad
Latencia medida en todas las ejecuciones de benchmark. P50 (mediana) y P95 (percentil 95) dan una imagen realista de la velocidad de respuesta bajo carga normal y máxima.
Tokens por segundo
Rendimiento en tokens por segundo, derivado de la latencia P50 medida. Más alto es mejor; las fluctuaciones reflejan la carga del lado del proveedor.
—
Capacidades
Disponibilidad
Disponibilidad
Sin datos todavía
Aún no hemos registrado suficientes llamadas a la API para mostrar estadísticas de disponibilidad de este modelo. Los datos aparecen una vez que el modelo comienza a recibir tráfico en vivo.
Veredictos del benchmark Tokonomix
FLUX.1 Kontext [max] adds image editing to multi-image fusion toolkit
FLUX.1 Kontext [max] has expanded its capabilities with the addition of image editing functionality, complementing its existing multi-image fusion features. This enhancement allows users to both combine multiple images and perform editing operations within the same workflow. The core multi-image fusion performance appears stable, maintaining the model's established approach to blending and compositing visual elements. The new image editing capability broadens the model's utility for creative workflows that require both image manipulation and fusion in a single pipeline. Users working on complex visual projects may find value in this consolidated approach, reducing the need to switch between separate tools for editing and fusion tasks. However, without comparative performance metrics between benchmark windows, it's difficult to assess whether the addition of editing features has impacted fusion quality or processing characteristics. The model continues to serve its primary function of multi-image fusion while now offering extended editing capabilities. Users should evaluate whether the integrated editing features meet their specific requirements compared to specialized single-purpose tools.
Calidad
—
Latencia p50
—
Ejecuciones de test
0
FLUX.1 Kontext [max] — Multi-Image Fusion
por fal.ai
- Ventana de contexto
- — tokens
- Precio de entrada
- — / 1M
- Precio de salida
- — / 1M
- Tier
- Tier A — Frontera
- Modalidad
- Texto
- Tipo de API
- REST · streaming
- Ejecuciones benchmark
- 187
Más de fal.ai