plumb through "generate audio" (only veo supports this??)

crates/desktop/artcraft/src/core/commands/enqueue/image_to_video/enqueue_image_to_video_command.rs

@@ -79,6 +79,11 @@ pub struct EnqueueImageToVideoRequest {
   /// Text prompt used to direct the video.
   pub prompt: Option<String>,
 
+  /// Optional.
+  /// Generate the video with audio, if there is an option.
+  /// Typically, this costs money when presented as an option, so we default to off.
+  pub generate_audio: Option<bool>,
+
   /// OPTIONAL.
   /// Only for Sora2 model currently.
   pub sora_orientation: Option<SoraOrientation>,

crates/desktop/artcraft/src/core/commands/enqueue/image_to_video/generic/handle_video_artcraft.rs

@@ -222,7 +222,7 @@ pub async fn handle_video_artcraft(
         prompt: request.prompt.clone(),
         duration: None,
         resolution: None,
-        generate_audio: None,
+        generate_audio: request.generate_audio,
       };
       let result = generate_veo_3_image_to_video(
         &app_env_configs.storyteller_host,
@@ -249,7 +249,7 @@ pub async fn handle_video_artcraft(
         prompt: request.prompt.clone(),
         duration: None,
         resolution: None,
-        generate_audio: None,
+        generate_audio: request.generate_audio,
       };
       let result = generate_veo_3_fast_image_to_video(
         &app_env_configs.storyteller_host,

frontend/libs/components/promptbox/src/lib/PromptBoxVideo.tsx

@@ -280,19 +280,15 @@ export const PromptBoxVideo = ({
       frontend_subscriber_id: subscriberId,
     };
 
+    if (selectedModel.generateWithSound) {
+      request.generate_audio = !!generateWithSound;
+    }
+
     if (selectedModel?.tauriId === "sora_2") {
       request.sora_orientation =
         resolution === "720p" ? "landscape" : "portrait";
     }
 
-    if (selectedModel?.tauriId === "veo_2") {
-      request.generate_with_sound = !!generateWithSound;
-    }
-
-    if (selectedModel?.tauriId === "veo_3") {
-      request.generate_with_sound = !!generateWithSound;
-    }
-
     await EnqueueImageToVideo(request);
 
     onEnqueuePressed?.(prompt, subscriberId);

frontend/libs/model-list/src/lib/classes/VideoModel.ts

@@ -47,6 +47,6 @@ export class VideoModel extends Model {
     this.endFrame = args.endFrame;
     this.requiresImage = args.requiresImage;
     this.sizeOptions = args.sizeOptions ?? [];
-    this.generateWithSound = args.generateWithSound;
+    this.generateWithSound = args.generateWithSound || false;
   }
 }

frontend/libs/model-list/src/lib/lists/VideoModels.ts

@@ -30,7 +30,7 @@ export const VIDEO_MODELS: VideoModel[] = [
     endFrame: true,
     requiresImage: true,
     progressBarTime: 100000,
-    generateWithSound: true,
+    generateWithSound: false,
   }),
   new VideoModel({
     id: "kling_2_1_pro",
@@ -45,7 +45,7 @@ export const VIDEO_MODELS: VideoModel[] = [
     endFrame: true,
     requiresImage: true,
     progressBarTime: 100000,
-    generateWithSound: true,
+    generateWithSound: false,
   }),
   new VideoModel({
     id: "kling_2_1_master",
@@ -60,7 +60,7 @@ export const VIDEO_MODELS: VideoModel[] = [
     endFrame: false,
     requiresImage: true,
     progressBarTime: 100000,
-    generateWithSound: true,
+    generateWithSound: false,
   }),
   new VideoModel({
     id: "seedance_1_0_lite",
@@ -114,7 +114,7 @@ export const VIDEO_MODELS: VideoModel[] = [
     startFrame: true,
     endFrame: false,
     requiresImage: true,
-    generateWithSound: true,
+    generateWithSound: false,
     progressBarTime: 100000,
   }),
   new VideoModel({

frontend/libs/tauri-api/src/lib/enqueue/EnqueueImageToVideo.ts

@@ -36,7 +36,7 @@ export interface EnqueueImageToVideoRequest {
   sora_orientation?: "portrait" | "landscape";
 
   // Optional. Whether to generate audio alongside the video (used by some models like Veo2)
-  generate_with_sound?: boolean;
+  generate_audio?: boolean;
 }
 
 interface RawEnqueueImageToVideoRequest {
@@ -47,7 +47,7 @@ interface RawEnqueueImageToVideoRequest {
   frontend_caller?: string;
   frontend_subscriber_id?: string;
   sora_orientation?: "portrait" | "landscape";
-  generate_with_sound?: boolean;
+  generate_audio?: boolean;
 }
 
 export interface EnqueueImageToVideoError extends CommandResult {
@@ -94,8 +94,8 @@ export const EnqueueImageToVideo = async (
     mutableRequest.sora_orientation = request.sora_orientation;
   }
 
-  if (typeof request.generate_with_sound === "boolean") {
-    mutableRequest.generate_with_sound = request.generate_with_sound;
+  if (typeof request.generate_audio === "boolean") {
+    mutableRequest.generate_audio = request.generate_audio;
   }
 
   const result = await invoke("enqueue_image_to_video_command", {