Mistral Large 4 1T-A52B (GGUF Q4_K_M + vision)
llama.cpp GGUF build (Q4_K_M, ggml-org) of Mistral Large 4, 1T MoE with 52B active; the F16 projector reads images.
Mistral Large 4 1T-A52B (GGUF Q4_K_M + vision)
Licence
Mistral Ecosystem Open-Weight License (a modified MIT licence). Copying, changing and redistributing the weights are allowed with the notice kept. Whoever operates a Model as a Service business - gives third parties access to the model for inference over an API or a hosted endpoint, with control over the inputs - must sign a separate commercial licence with Mistral before any commercial use. A neuronet serves inference to its clients, so the catalogue marks commercial use as not granted: a neuronet owner needs Mistral's commercial licence first.
Miner
The chat template of this build carries the string literal 'sentinel'. It needs a miner that accepts that literal; the 0.0.10 release refuses any '__' in a chat template.