preview

Preview

Fast, lightweight model for text and vision tasks.
duce

Duce

High-performance multimodal with SOTA music generation and Mixture of Models (MoM).
magma

Magma

Maximum capability with 1M context, SOTA video, and Mixture of Models (MoM).

​
Preview

The fastest model layer, optimized for text and vision tasks. Capabilities:
  • Function calling
  • Structured output
  • Reasoning
Best for: Quick text completions, image analysis, structured data extraction, and function calling where speed matters most.

​
Duce

High-performance multimodal model with state-of-the-art music generation and broad media capabilities. Capabilities:
  • SOTA music generation
  • Function calling
  • Reasoning
  • Mixture of Models (MoM)
  • Text-to-Speech (TTS)
  • Image-to-Image
  • Text/Image-to-Video
  • Video-to-Video
  • Text-to-Music / Music-to-Music
Best for: Music generation, multimodal content creation, and complex reasoning tasks requiring media output.

​
Magma

The most capable model layer with the largest context window and state-of-the-art video generation. Capabilities:
  • SOTA video generation
  • Function calling
  • Reasoning
  • Mixture of Models (MoM)
  • Text-to-Speech (TTS)
  • Image-to-Image
  • Text/Image-to-Video
  • Video-to-Video
  • Text-to-Music / Music-to-Music
Best for: Long-context tasks, video generation, complex agentic workflows, and maximum quality output across all modalities.

​
Comparison

​
Mixture of Models (MoM)

Duce and Magma support Mixture of Models (MoM) — a technique that routes requests across multiple specialized models within the layer to produce the best output. MoM is automatically applied when using these model layers; no additional configuration is required.