diff --git a/whisper.cpp b/whisper.cpp
index f8c1ada..3633cdb 100644
--- a/whisper.cpp
+++ b/whisper.cpp
@@ -2591,7 +2591,7 @@ int whisper_full(
     }
 
     // Prepend the prompt tokens to the prompt_past
-    if (params.prompt_tokens) {
+    if (params.prompt_tokens && params.prompt_n_tokens > 0) {
         // Parse tokens from the pointer (it points to an std::vector)
         for (int i = 0; i < params.prompt_n_tokens; i++) {
             prompt_past.push_back(params.prompt_tokens[i]);
diff --git a/whisper.h b/whisper.h
index ad989e9..1b2a042 100644
--- a/whisper.h
+++ b/whisper.h
@@ -3,7 +3,6 @@
 
 #include <stdint.h>
 #include <stdbool.h>
-#include <vector>
 
 #ifdef WHISPER_SHARED
 #    ifdef _WIN32
@@ -209,7 +208,7 @@ extern "C" {
         bool speed_up;  // speed-up the audio by 2x using Phase Vocoder
         int  audio_ctx; // overwrite the audio context size (0 = use default)
 
-        // std::vector<whisper_token>: tokents to provide the whisper model as initial prompt
+        // std::vector<whisper_token>: tokens to provide the whisper model as initial prompt
         const whisper_token * prompt_tokens;
         int prompt_n_tokens;