diff --git a/backend/cpp/llama-cpp/Makefile b/backend/cpp/llama-cpp/Makefile
index 1755caaae..4d6d1c27a 100644
--- a/backend/cpp/llama-cpp/Makefile
+++ b/backend/cpp/llama-cpp/Makefile
@@ -1,5 +1,5 @@
 
-LLAMA_VERSION?=2b4cbd2834e427024bc7f935a1f232aecac6679b
+LLAMA_VERSION?=f6b533d898ce84bae8d9fa8dfc6697ac087800bf
 LLAMA_REPO?=https://github.com/ggerganov/llama.cpp
 
 CMAKE_ARGS?=
diff --git a/backend/cpp/llama-cpp/grpc-server.cpp b/backend/cpp/llama-cpp/grpc-server.cpp
index 0f29cc755..a9870e242 100644
--- a/backend/cpp/llama-cpp/grpc-server.cpp
+++ b/backend/cpp/llama-cpp/grpc-server.cpp
@@ -778,8 +778,8 @@ public:
             if (!params.mmproj.path.empty()) {
                 error_msg += " (with mmproj: " + params.mmproj.path + ")";
             }
-            if (params.has_speculative() && !params.speculative.model.path.empty()) {
-                error_msg += " (with draft model: " + params.speculative.model.path + ")";
+            if (params.speculative.has_dft() && !params.speculative.mparams_dft.path.empty()) {
+                error_msg += " (with draft model: " + params.speculative.mparams_dft.path + ")";
             }
             
             // Add captured error details if available