← Back to Models
Mistral AImoe
Mixtral 8x22B
176B total / 44B active · load estimate uses total • 107.1GB VRAM (Q4 estimate) • 65,536 context
Estimate · Q4_K_M · 4k context · batch 1. For dense 7B–70B this sits ~5–8% above the Q4_K_M GGUF file. Not peak runtime VRAM and not a measured bench.
Specifications
Parameters
176B total / 44B active · load estimate uses total
VRAM (Q4 est.)
107.1 GB
VRAM (FP16)
352 GB
Context Window
65,536
Architecture
moe
License
Apache-2.0
Finding cloud alternatives...
Run Mixtral 8x22B on Radeon Instinct MI250X
~107.1GB VRAM needed at Q4. Radeon Instinct MI250X has 128GB — buy hardware or rent cloud.
We may earn a commission from cloud and hardware partners at no extra cost to you.
Find the cheapest GPU that can run Mixtral 8x22B
Compatibility Lab — check every GPU × quantization combination