What Models?

Pick the right model for your GPU — in seconds.

What GPU do you need to run Muse Spark Open 816B A42B?

822.04B params 1M max context VisionTool useReasoning

The smallest GPU memory that runs Muse Spark Open 816B A42B well is 512 GB (Q4_K_M, e.g. M3 Ultra). 2 of 132 GPUs run it well. It's a mixture-of-experts model, so all weights must fit in memory but speed depends only on the active experts.

VRAM needed per quantization

QuantizationWeights16K context32K contextFull 1M
Q4_K_M465.3 GB466.3 GB467.2 GB527.7 GB
Q8_0811.8 GB812.8 GB813.7 GB874.3 GB

Weights plus fp16 KV cache. Leave ~1 GB headroom for the inference engine.

GPUs that can run Muse Spark Open 816B A42B

Best quantization that fits on each GPU, with the context and speed it gets. For chips with several memory sizes, the smallest that works is shown.

Apple

  • M3 Ultra 512 GB Q4_K_M 765K context · ~33 tok/s
  • M5 Ultra 512 GB Q4_K_M 765K context · ~48 tok/s
Check your exact setup in the full tool →