diff --git a/src/modules/aix/client/aix.client.chatGenerateRequest.ts b/src/modules/aix/client/aix.client.chatGenerateRequest.ts
index 2b7ec4b4d..fd1f0ee3b 100644
--- a/src/modules/aix/client/aix.client.chatGenerateRequest.ts
+++ b/src/modules/aix/client/aix.client.chatGenerateRequest.ts
@@ -137,6 +137,11 @@ export async function aixCGR_ChatSequence_FromDMessagesOrThrow(
             uMsg.parts.push(uFragment.part);
             break;
 
+          case 'audio_ref':
+            // Note: unsupported for now (shall await for the audio blob to upload it as inline audio for the model)
+            console.warn('aixCGR_FromDMessages: audio_ref part from User not implemented yet');
+            break;
+
           case 'image_ref':
             // note, we don't resize, as the user image is resized following the user's preferences
             try {
@@ -210,8 +215,7 @@ export async function aixCGR_ChatSequence_FromDMessagesOrThrow(
             break;
 
           case 'doc':
-            // TODO
-            console.warn('aixCGR_FromDMessages: doc part from Assistant not implemented yet');
+            console.warn('aixCGR_FromDMessages: doc part from Assistant not possible');
             // mMsg.parts.push(aFragment.part);
             break;
 
@@ -220,6 +224,11 @@ export async function aixCGR_ChatSequence_FromDMessagesOrThrow(
             modelMessage.parts.push({ pt: 'text', text: `[ERROR] ${aFragment.part.error}` });
             break;
 
+          case 'audio_ref':
+            // Note: unsupported for now (shall await for the audio blob to upload it as inline audio for the model)
+            console.warn('aixCGR_FromDMessages: audio_ref part from Assistant not implemented yet');
+            break;
+
           case 'image_ref':
             // TODO: rescale shall be dependent on the LLM here - and be careful with the high-res options, as they can
             //  be really space consuming. how to choose between high and low? global option?