BahamutRU/DeepSeek-V4-Flash-Vision-Exp-Q8_0-MXFP4-Q3_K-Q2_K-mixed-GGUF
Not comparable yet vs deepseek-ai/DeepSeek-V4-Flash-Vision-Exp. Not verifiable yet: the GGUF's tensors don't map onto the other model's names (88 of 72633/1328 match), usually an architecture llama.cpp's name map doesn't cover yet.