diff --git a/common/arg.cpp b/common/arg.cpp
index b5741c859..9b08b9a65 100644
--- a/common/arg.cpp
+++ b/common/arg.cpp
@@ -2120,13 +2120,6 @@ common_params_context common_params_parser_init(common_params & params, llama_ex
             params.mmproj.url = value;
         }
     ).set_examples({LLAMA_EXAMPLE_LLAVA}));
-    add_opt(common_arg(
-        {"--no-mmproj"},
-        "explicitly disable multimodal projector, useful when using -hf",
-        [](common_params & params) {
-            params.no_mmproj = true;
-        }
-    ).set_examples({LLAMA_EXAMPLE_LLAVA}));
     add_opt(common_arg(
         {"--image"}, "FILE",
         "path to an image file. use with multimodal models. Specify multiple times for batching",
diff --git a/common/common.h b/common/common.h
index 70d3ef8f2..e6eaa8e80 100644
--- a/common/common.h
+++ b/common/common.h
@@ -342,7 +342,6 @@ struct common_params {
 
     // multimodal models (see examples/llava)
     struct common_params_model mmproj;
-    bool no_mmproj = false;         // explicitly disable multimodal model
     std::vector<std::string> image; // path to image file(s)
 
     // embedding