mulmocast 2.6.7 → 2.6.9

package/lib/actions/image_references.js

@@ -121,6 +121,7 @@ const generateReferenceMovie = async (inputs) => {
                 params: {
                     model: movieAgentInfo.movieParams.model,
                     canvasSize: context.presentationStyle.canvasSize,
+                    generateAudio: movieAgentInfo.movieParams.generateAudio,
                 },
             },
         },

package/lib/actions/images.d.ts

@@ -324,6 +324,7 @@ export declare const beat_graph_data: {
                     canvasSize: string;
                     vertexai_project: string;
                     vertexai_location: string;
+                    generateAudio: string;
                 };
             };
             defaultValue: {};

package/lib/actions/images.js

@@ -194,6 +194,7 @@ export const beat_graph_data = {
                     canvasSize: ":context.presentationStyle.canvasSize",
                     vertexai_project: ":preprocessor.movieAgentInfo.movieParams.vertexai_project",
                     vertexai_location: ":preprocessor.movieAgentInfo.movieParams.vertexai_location",
+                    generateAudio: ":preprocessor.movieAgentInfo.movieParams.generateAudio",
                 },
             },
             defaultValue: {},

package/lib/agents/movie_genai_agent.js

@@ -1,10 +1,10 @@
 import { readFileSync, writeFileSync } from "fs";
 import { GraphAILogger, sleep } from "graphai";
 import { GoogleGenAI, PersonGeneration } from "@google/genai";
-import { apiKeyMissingError, agentGenerationError, agentInvalidResponseError, imageAction, movieFileTarget, videoDurationTarget, hasCause, } from "../utils/error_cause.js";
+import { apiKeyMissingError, agentGenerationError, agentInvalidResponseError, imageAction, movieFileTarget, videoDurationTarget, unsupportedModelTarget, hasCause, } from "../utils/error_cause.js";
 import { getAspectRatio } from "../utils/utils.js";
 import { ASPECT_RATIOS } from "../types/const.js";
-import { getModelDuration, provider2MovieAgent } from "../types/provider2agent.js";
+import { getModelDuration, provider2MovieAgent, AUDIO_MODE_NEVER, AUDIO_MODE_ALWAYS } from "../types/provider2agent.js";
 const pollUntilDone = async (ai, operation) => {
     const response = { operation };
     while (!response.operation.done) {
@@ -154,6 +154,18 @@ export const movieGenAIAgent = async ({ namedInputs, params, config, }) => {
                 cause: agentGenerationError("movieGenAIAgent", imageAction, videoDurationTarget),
             });
         }
+        // Check generateAudio compatibility (Google API has no toggle)
+        if (params.generateAudio !== undefined) {
+            const audio = provider2MovieAgent.google.modelParams[model]?.audio ?? { mode: AUDIO_MODE_NEVER };
+            if (audio.mode === AUDIO_MODE_NEVER && params.generateAudio === true) {
+                throw new Error(`Model ${model} does not support audio generation`, {
+                    cause: agentGenerationError("movieGenAIAgent", imageAction, unsupportedModelTarget),
+                });
+            }
+            else if (audio.mode === AUDIO_MODE_ALWAYS && params.generateAudio === false) {
+                GraphAILogger.warn(`movieGenAIAgent: model ${model} always generates audio — ignoring generateAudio=false`);
+            }
+        }
         const isVertexAI = !!params.vertexai_project;
         const ai = isVertexAI
             ? new GoogleGenAI({

package/lib/agents/movie_replicate_agent.js

@@ -1,8 +1,8 @@
 import { readFileSync } from "fs";
 import { GraphAILogger } from "graphai";
 import Replicate from "replicate";
-import { apiKeyMissingError, agentGenerationError, agentInvalidResponseError, imageAction, movieFileTarget, videoDurationTarget, unsupportedModelTarget, } from "../utils/error_cause.js";
-import { provider2MovieAgent, getModelDuration } from "../types/provider2agent.js";
+import { apiKeyMissingError, agentGenerationError, agentInvalidResponseError, hasCause, imageAction, movieFileTarget, videoDurationTarget, unsupportedModelTarget, } from "../utils/error_cause.js";
+import { provider2MovieAgent, getModelDuration, AUDIO_MODE_OPTIONAL, AUDIO_MODE_NEVER, AUDIO_MODE_ALWAYS } from "../types/provider2agent.js";
 function replicate_get_videoUrl(output) {
     if (typeof output === "string")
         return output;
@@ -10,7 +10,7 @@ function replicate_get_videoUrl(output) {
         return output.url();
     return undefined;
 }
-async function generateMovie(model, apiKey, prompt, imagePath, lastFrameImagePath, referenceImages, aspectRatio, duration) {
+async function generateMovie(model, apiKey, prompt, imagePath, lastFrameImagePath, referenceImages, aspectRatio, duration, generateAudio) {
     const replicate = new Replicate({
         auth: apiKey,
     });
@@ -77,6 +77,21 @@ async function generateMovie(model, apiKey, prompt, imagePath, lastFrameImagePat
             GraphAILogger.warn(`movieReplicateAgent: model ${model} does not support lastFrame — ignoring lastFrameImageName`);
         }
     }
+    // Add generate_audio if the model supports it
+    const audio = provider2MovieAgent.replicate.modelParams[model].audio;
+    if (generateAudio !== undefined) {
+        if (audio.mode === AUDIO_MODE_OPTIONAL) {
+            input[audio.param] = generateAudio;
+        }
+        else if (audio.mode === AUDIO_MODE_NEVER && generateAudio === true) {
+            throw new Error(`Model ${model} does not support audio generation`, {
+                cause: agentGenerationError("movieReplicateAgent", imageAction, unsupportedModelTarget),
+            });
+        }
+        else if (audio.mode === AUDIO_MODE_ALWAYS && generateAudio === false) {
+            GraphAILogger.warn(`movieReplicateAgent: model ${model} always generates audio — ignoring generateAudio=false`);
+        }
+    }
     try {
         const output = await replicate.run(model, { input });
         // Download the generated video
@@ -134,12 +149,15 @@ export const movieReplicateAgent = async ({ namedInputs, params, config, }) => {
         });
     }
     try {
-        const buffer = await generateMovie(model, apiKey, prompt, imagePath, lastFrameImagePath, referenceImages, aspectRatio, duration);
+        const buffer = await generateMovie(model, apiKey, prompt, imagePath, lastFrameImagePath, referenceImages, aspectRatio, duration, params.generateAudio);
         if (buffer) {
             return { buffer };
         }
     }
     catch (error) {
+        if (hasCause(error)) {
+            throw error;
+        }
         GraphAILogger.info("Failed to generate movie:", error.message);
     }
     throw new Error("ERROR: generateMovie returned undefined", {

package/lib/agents/tts_google_agent.js

@@ -2,6 +2,10 @@ import { GraphAILogger } from "graphai";
 import * as textToSpeech from "@google-cloud/text-to-speech";
 import { agentGenerationError, audioAction, audioFileTarget } from "../utils/error_cause.js";
 const client = new textToSpeech.TextToSpeechClient();
+// Hard cap so a hung Google TTS RPC can't pin a beat indefinitely.
+// Most synthesizeSpeech calls return in seconds; 60s leaves headroom
+// for long inputs and slow regions while still failing loud.
+const SYNTHESIZE_TIMEOUT_MS = 60_000;
 const getPrompt = (text, instructions) => {
     if (instructions) {
         return `### DIRECTOR'S NOTES\n${instructions}\n\n#### TRANSCRIPT\n${text}`;
@@ -37,7 +41,7 @@ export const ttsGoogleAgent = async ({ namedInputs, params }) => {
     };
     try {
         // Call the Text-to-Speech API
-        const [response] = await client.synthesizeSpeech(request);
+        const [response] = await client.synthesizeSpeech(request, { timeout: SYNTHESIZE_TIMEOUT_MS });
         return { buffer: response.audioContent };
     }
     catch (e) {
@@ -47,11 +51,23 @@ export const ttsGoogleAgent = async ({ namedInputs, params }) => {
             };
         }
         GraphAILogger.info(e);
-        throw new Error("TTS Google Error", {
+        // gRPC errors from @google-cloud/text-to-speech are ServiceError
+        // (extends Error with a `details` string). Surface that human-readable
+        // text so callers don't see only "TTS Google Error".
+        throw new Error(`TTS Google Error: ${grpcErrorDetail(e)}`, {
             cause: agentGenerationError("ttsGoogleAgent", audioAction, audioFileTarget),
         });
     }
 };
+const grpcErrorDetail = (e) => {
+    if (e instanceof Error) {
+        const details = e.details;
+        if (typeof details === "string" && details)
+            return details;
+        return e.message;
+    }
+    return String(e);
+};
 const ttsGoogleAgentInfo = {
     name: "ttsGoogleAgent",
     agent: ttsGoogleAgent,

package/lib/methods/mulmo_presentation_style.d.ts

@@ -180,6 +180,7 @@ export declare const MulmoPresentationStyleMethods: {
                 referenceType: "ASSET" | "STYLE";
             }[] | undefined;
             concurrency?: number | undefined;
+            generateAudio?: boolean | undefined;
             speed?: number | undefined;
         };
         keyName: string;

package/lib/types/agent.d.ts

@@ -84,6 +84,7 @@ export type GoogleMovieAgentParams = ImageAgentParams & {
     duration?: number;
     vertexai_project?: string;
     vertexai_location?: string;
+    generateAudio?: boolean;
 };
 export type ReplicateMovieAgentParams = {
     model: `${string}/${string}` | undefined;
@@ -92,6 +93,7 @@ export type ReplicateMovieAgentParams = {
         height: number;
     };
     duration?: number;
+    generateAudio?: boolean;
 };
 export type ReplicateSoundEffectAgentParams = {
     model: `${string}/${string}` | undefined;

package/lib/types/provider2agent.d.ts

@@ -70,32 +70,47 @@ export declare const provider2ImageAgent: {
     };
 };
 export type ReplicateModel = `${string}/${string}`;
+export declare const AUDIO_MODE_NEVER: "never";
+export declare const AUDIO_MODE_ALWAYS: "always";
+export declare const AUDIO_MODE_OPTIONAL: "optional";
+type MovieAudioSpec = {
+    mode: typeof AUDIO_MODE_NEVER;
+} | {
+    mode: typeof AUDIO_MODE_ALWAYS;
+} | {
+    mode: typeof AUDIO_MODE_OPTIONAL;
+    param: string;
+};
+type ReplicateMovieModelParams = {
+    durations: number[];
+    start_image: string | undefined;
+    last_image?: string;
+    reference_images_param?: string;
+    audio: MovieAudioSpec;
+    price_per_sec: number;
+};
+type GoogleMovieModelParams = {
+    durations: number[];
+    supportsDuration: boolean;
+    supportsLastFrame: boolean;
+    supportsReferenceImages: boolean;
+    supportsPersonGeneration: boolean;
+    audio: MovieAudioSpec;
+};
 export declare const provider2MovieAgent: {
     replicate: {
         agentName: string;
         defaultModel: ReplicateModel;
         keyName: string;
         models: string[];
-        modelParams: Record<ReplicateModel, {
-            durations: number[];
-            start_image: string | undefined;
-            last_image?: string;
-            reference_images_param?: string;
-            price_per_sec: number;
-        }>;
+        modelParams: Record<ReplicateModel, ReplicateMovieModelParams>;
     };
     google: {
         agentName: string;
         defaultModel: string;
         models: string[];
         keyName: string;
-        modelParams: Record<string, {
-            durations: number[];
-            supportsDuration: boolean;
-            supportsLastFrame: boolean;
-            supportsReferenceImages: boolean;
-            supportsPersonGeneration: boolean;
-        }>;
+        modelParams: Record<string, GoogleMovieModelParams>;
     };
     mock: {
         agentName: string;
@@ -182,3 +197,4 @@ export declare const llm: (keyof typeof provider2LLMAgent)[];
 export type LLM = keyof typeof provider2LLMAgent;
 export declare const htmlLLMProvider: string[];
 export declare const getModelDuration: (provider: keyof typeof provider2MovieAgent, model: string, movieDuration?: number) => number | undefined;
+export {};

package/lib/types/provider2agent.js

@@ -44,7 +44,7 @@ export const provider2TTSAgent = {
         models: ["mock-model"],
     },
 };
-export const gptImages = ["gpt-image-1.5", "gpt-image-1", "gpt-image-1-mini"];
+export const gptImages = ["gpt-image-2", "gpt-image-1.5", "gpt-image-1", "gpt-image-1-mini"];
 export const provider2ImageAgent = {
     openai: {
         agentName: "imageOpenaiAgent",
@@ -79,6 +79,9 @@ export const provider2ImageAgent = {
         keyName: "",
     },
 };
+export const AUDIO_MODE_NEVER = "never";
+export const AUDIO_MODE_ALWAYS = "always";
+export const AUDIO_MODE_OPTIONAL = "optional";
 export const provider2MovieAgent = {
     replicate: {
         agentName: "movieReplicateAgent",
@@ -87,6 +90,8 @@ export const provider2MovieAgent = {
         models: [
             "bytedance/seedance-1-lite",
             "bytedance/seedance-1-pro",
+            "bytedance/seedance-2.0",
+            "bytedance/seedance-2.0-fast",
             "kwaivgi/kling-v1.6-pro",
             "kwaivgi/kling-v2.1",
             "kwaivgi/kling-v2.1-master",
@@ -113,37 +118,58 @@ export const provider2MovieAgent = {
                 durations: [5, 10],
                 start_image: "image",
                 last_image: "last_frame_image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.036, // in USD
             },
             "bytedance/seedance-1-pro": {
                 durations: [5, 10],
                 start_image: "image",
                 last_image: "last_frame_image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.15,
             },
+            "bytedance/seedance-2.0": {
+                durations: [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15],
+                start_image: "image",
+                last_image: "last_frame_image",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
+                price_per_sec: 0.29,
+            },
+            "bytedance/seedance-2.0-fast": {
+                durations: [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15],
+                start_image: "image",
+                last_image: "last_frame_image",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
+                price_per_sec: 0.22,
+            },
             "kwaivgi/kling-v1.6-pro": {
                 durations: [5, 10],
                 start_image: "start_image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.095,
             },
             "kwaivgi/kling-v2.1": {
                 durations: [5, 10],
                 start_image: "start_image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.05,
             },
             "kwaivgi/kling-v2.1-master": {
                 durations: [5, 10],
                 start_image: "start_image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.28,
             },
             "google/veo-2": {
                 durations: [5, 6, 7, 8],
                 start_image: "image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.5,
             },
             "google/veo-3": {
                 durations: [8],
                 start_image: "image",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
                 price_per_sec: 0.75,
             },
             "google/veo-3.1": {
@@ -151,71 +177,84 @@ export const provider2MovieAgent = {
                 start_image: "image",
                 last_image: "last_frame_image",
                 reference_images_param: "reference_images",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
                 price_per_sec: 0.75,
             },
             "google/veo-3.1-fast": {
                 durations: [4, 6, 8],
                 start_image: "image",
                 last_image: "last_frame_image",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
                 price_per_sec: 0.4,
             },
             "google/veo-3.1-lite": {
                 durations: [4, 6, 8],
                 start_image: "image",
                 last_image: "last_frame",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.05,
             },
             "google/veo-3-fast": {
                 durations: [8],
                 start_image: "image",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
                 price_per_sec: 0.4,
             },
             "minimax/video-01": {
                 durations: [6],
                 start_image: "first_frame_image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.5,
             },
             "minimax/hailuo-02": {
                 durations: [6], // NOTE: 10 for only 720p
                 start_image: "first_frame_image",
                 last_image: "end_image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.08,
             },
             "minimax/hailuo-02-fast": {
                 durations: [6, 10], // NOTE: 512P
                 start_image: "first_frame_image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.0166,
             },
             "pixverse/pixverse-v4.5": {
                 durations: [5, 8],
                 start_image: "image",
                 last_image: "last_frame_image",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "sound_effect_switch" },
                 price_per_sec: 0.12,
             },
             "wan-video/wan-2.2-i2v-fast": {
                 durations: [5],
                 start_image: "image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.012,
             },
             "wan-video/wan-2.2-t2v-fast": {
                 durations: [5],
                 start_image: undefined,
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.012,
             },
             "xai/grok-imagine-video": {
                 durations: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15],
                 start_image: "image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.08,
             },
             "xai/grok-imagine-r2v": {
                 durations: [1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
                 start_image: undefined,
                 reference_images_param: "reference_images",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.08,
             },
             "runwayml/gen-4.5": {
                 durations: [5, 10],
                 start_image: "image",
+                audio: { mode: AUDIO_MODE_NEVER },
                 price_per_sec: 0.25,
             },
             "kwaivgi/kling-v3-omni-video": {
@@ -223,6 +262,7 @@ export const provider2MovieAgent = {
                 start_image: "start_image",
                 last_image: "end_image",
                 reference_images_param: "reference_images",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
                 price_per_sec: 0.3,
             },
             "kwaivgi/kling-v3-video": {
@@ -230,6 +270,7 @@ export const provider2MovieAgent = {
                 start_image: "start_image",
                 last_image: "end_image",
                 reference_images_param: "reference_images",
+                audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
                 price_per_sec: 0.3,
             },
         },
@@ -246,6 +287,7 @@ export const provider2MovieAgent = {
                 supportsLastFrame: true,
                 supportsReferenceImages: false,
                 supportsPersonGeneration: false,
+                audio: { mode: AUDIO_MODE_ALWAYS },
             },
             "veo-3.1-generate-preview": {
                 durations: [4, 6, 8],
@@ -253,6 +295,7 @@ export const provider2MovieAgent = {
                 supportsLastFrame: true,
                 supportsReferenceImages: true,
                 supportsPersonGeneration: false,
+                audio: { mode: AUDIO_MODE_ALWAYS },
             },
             "veo-3.0-generate-001": {
                 durations: [8],
@@ -260,6 +303,7 @@ export const provider2MovieAgent = {
                 supportsLastFrame: false,
                 supportsReferenceImages: false,
                 supportsPersonGeneration: false,
+                audio: { mode: AUDIO_MODE_ALWAYS },
             },
             "veo-2.0-generate-001": {
                 durations: [5, 6, 8],
@@ -267,6 +311,7 @@ export const provider2MovieAgent = {
                 supportsLastFrame: false, // Vertex AI only
                 supportsReferenceImages: false,
                 supportsPersonGeneration: true,
+                audio: { mode: AUDIO_MODE_NEVER },
             },
         },
     },

package/lib/types/schema.d.ts

@@ -3691,6 +3691,7 @@ export declare const mulmoMovieParamsSchema: z.ZodObject<{
         }>;
     }, z.core.$strip>>>;
     concurrency: z.ZodOptional<z.ZodNumber>;
+    generateAudio: z.ZodOptional<z.ZodBoolean>;
 }, z.core.$strip>;
 export declare const mulmoBeatSchema: z.ZodObject<{
     speaker: z.ZodOptional<z.ZodString>;
@@ -6585,6 +6586,7 @@ export declare const mulmoBeatSchema: z.ZodObject<{
             }>;
         }, z.core.$strip>>>;
         concurrency: z.ZodOptional<z.ZodNumber>;
+        generateAudio: z.ZodOptional<z.ZodBoolean>;
         speed: z.ZodOptional<z.ZodNumber>;
     }, z.core.$strip>>;
     soundEffectParams: z.ZodOptional<z.ZodObject<{
@@ -7074,6 +7076,7 @@ export declare const mulmoPresentationStyleSchema: z.ZodObject<{
             }>;
         }, z.core.$strip>>>;
         concurrency: z.ZodOptional<z.ZodNumber>;
+        generateAudio: z.ZodOptional<z.ZodBoolean>;
     }, z.core.$strip>>>;
     soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
         provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{
@@ -7578,6 +7581,7 @@ export declare const mulmoScriptSchema: z.ZodObject<{
             }>;
         }, z.core.$strip>>>;
         concurrency: z.ZodOptional<z.ZodNumber>;
+        generateAudio: z.ZodOptional<z.ZodBoolean>;
     }, z.core.$strip>>>;
     soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
         provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{
@@ -10611,6 +10615,7 @@ export declare const mulmoScriptSchema: z.ZodObject<{
                 }>;
             }, z.core.$strip>>>;
             concurrency: z.ZodOptional<z.ZodNumber>;
+            generateAudio: z.ZodOptional<z.ZodBoolean>;
             speed: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strip>>;
         soundEffectParams: z.ZodOptional<z.ZodObject<{
@@ -11175,6 +11180,7 @@ export declare const mulmoStudioSchema: z.ZodObject<{
                 }>;
             }, z.core.$strip>>>;
             concurrency: z.ZodOptional<z.ZodNumber>;
+            generateAudio: z.ZodOptional<z.ZodBoolean>;
         }, z.core.$strip>>>;
         soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{
@@ -14208,6 +14214,7 @@ export declare const mulmoStudioSchema: z.ZodObject<{
                     }>;
                 }, z.core.$strip>>>;
                 concurrency: z.ZodOptional<z.ZodNumber>;
+                generateAudio: z.ZodOptional<z.ZodBoolean>;
                 speed: z.ZodOptional<z.ZodNumber>;
             }, z.core.$strip>>;
             soundEffectParams: z.ZodOptional<z.ZodObject<{
@@ -14708,6 +14715,7 @@ export declare const mulmoPromptTemplateSchema: z.ZodObject<{
                 }>;
             }, z.core.$strip>>>;
             concurrency: z.ZodOptional<z.ZodNumber>;
+            generateAudio: z.ZodOptional<z.ZodBoolean>;
         }, z.core.$strip>>>;
         soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{
@@ -15206,6 +15214,7 @@ export declare const mulmoPromptTemplateFileSchema: z.ZodObject<{
                 }>;
             }, z.core.$strip>>>;
             concurrency: z.ZodOptional<z.ZodNumber>;
+            generateAudio: z.ZodOptional<z.ZodBoolean>;
         }, z.core.$strip>>>;
         soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{

package/lib/types/schema.js

@@ -499,6 +499,7 @@ export const mulmoMovieParamsSchema = z.object({
         .optional()
         .describe("Style/asset reference images (Veo 3.1). Mutually exclusive with imageName/lastFrameImageName"),
     concurrency: z.number().int().positive().optional().describe("Max concurrent movie generation requests"),
+    generateAudio: z.boolean().optional().describe("Request audio generation in the video (model-dependent)"),
 });
 export const mulmoBeatSchema = z
     .object({

package/lib/utils/context.d.ts

@@ -282,6 +282,7 @@ export declare const createStudioData: (_mulmoScript: MulmoScript, fileName: str
                 referenceType: "ASSET" | "STYLE";
             }[] | undefined;
             concurrency?: number | undefined;
+            generateAudio?: boolean | undefined;
         };
         soundEffectParams: {
             provider?: string | undefined;
@@ -2022,6 +2023,7 @@ export declare const createStudioData: (_mulmoScript: MulmoScript, fileName: str
                     referenceType: "ASSET" | "STYLE";
                 }[] | undefined;
                 concurrency?: number | undefined;
+                generateAudio?: boolean | undefined;
                 speed?: number | undefined;
             } | undefined;
             soundEffectParams?: {
@@ -2518,6 +2520,7 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                     referenceType: "ASSET" | "STYLE";
                 }[] | undefined;
                 concurrency?: number | undefined;
+                generateAudio?: boolean | undefined;
             };
             soundEffectParams: {
                 provider?: string | undefined;
@@ -4258,6 +4261,7 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                         referenceType: "ASSET" | "STYLE";
                     }[] | undefined;
                     concurrency?: number | undefined;
+                    generateAudio?: boolean | undefined;
                     speed?: number | undefined;
                 } | undefined;
                 soundEffectParams?: {
@@ -4761,6 +4765,7 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                 referenceType: "ASSET" | "STYLE";
             }[] | undefined;
             concurrency?: number | undefined;
+            generateAudio?: boolean | undefined;
         };
         soundEffectParams: {
             provider?: string | undefined;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mulmocast",
-  "version": "2.6.7",
+  "version": "2.6.9",
   "description": "",
   "type": "module",
   "main": "lib/index.node.js",
@@ -89,7 +89,7 @@
   "homepage": "https://github.com/receptron/mulmocast-cli#readme",
   "dependencies": {
     "@google-cloud/text-to-speech": "^6.4.0",
-    "@google/genai": "^1.49.0",
+    "@google/genai": "^1.50.1",
     "@graphai/anthropic_agent": "^2.0.12",
     "@graphai/browserless_agent": "^2.0.2",
     "@graphai/gemini_agent": "^2.0.5",
@@ -99,8 +99,8 @@
     "@graphai/stream_agent_filter": "^2.0.3",
     "@graphai/vanilla": "^2.0.12",
     "@graphai/vanilla_node_agents": "^2.0.4",
-    "@inquirer/input": "^5.0.11",
-    "@inquirer/select": "^5.1.3",
+    "@inquirer/input": "^5.0.12",
+    "@inquirer/select": "^5.1.4",
     "@modelcontextprotocol/sdk": "^1.29.0",
     "@mozilla/readability": "^0.6.0",
     "@tavily/core": "^0.5.11",
@@ -109,11 +109,11 @@
     "dotenv": "^17.4.2",
     "fluent-ffmpeg": "^2.1.3",
     "graphai": "^2.0.16",
-    "jsdom": "^29.0.2",
-    "marked": "^18.0.0",
+    "jsdom": "^29.1.0",
+    "marked": "^18.0.2",
     "mulmocast-vision": "^1.0.9",
-    "ora": "^9.3.0",
-    "puppeteer": "^24.40.0",
+    "ora": "^9.4.0",
+    "puppeteer": "^24.42.0",
     "replicate": "^1.4.0",
     "yaml": "^2.8.3",
     "yargs": "^18.0.0",
@@ -127,16 +127,16 @@
     "@types/jsdom": "^28.0.1",
     "@types/yargs": "^17.0.35",
     "cross-env": "^10.1.0",
-    "eslint": "^10.2.0",
+    "eslint": "^10.2.1",
     "eslint-config-prettier": "^10.1.8",
     "eslint-plugin-import": "^2.32.0",
     "eslint-plugin-prettier": "^5.5.5",
-    "eslint-plugin-sonarjs": "^4.0.2",
+    "eslint-plugin-sonarjs": "^4.0.3",
     "globals": "^17.5.0",
-    "prettier": "^3.8.2",
+    "prettier": "^3.8.3",
     "tsx": "^4.21.0",
-    "typescript": "6.0.2",
-    "typescript-eslint": "^8.58.1"
+    "typescript": "6.0.3",
+    "typescript-eslint": "^8.59.1"
   },
   "engines": {
     "node": ">=22.0.0"

package/scripts/test/README.md

@@ -106,7 +106,7 @@ Language setting tests
 Provider-specific feature tests
 
 - [**test_hello_google.json**](./test_hello_google.json) - Google TTS専用テスト / Google TTS specific test
-- [**gpt.json**](./gpt.json) - GPTモデルテスト / GPT model test
+- [**test_gpt_image.json**](./test_gpt_image.json) - GPT image model test
 - [**mulmo_story.json**](./mulmo_story.json) - ストーリー形式テスト / Story format test
 - [**nano_banana.json**](./nano_banana.json) - カスタムサンプル / Custom sample
 

package/scripts/test/test_generate_audio.json

@@ -0,0 +1,33 @@
+{
+  "$mulmocast": { "version": "1.1" },
+  "title": "generateAudio Test",
+  "lang": "en",
+  "canvasSize": { "width": 1280, "height": 720 },
+  "audioParams": {
+    "bgmVolume": 0,
+    "suppressSpeech": true
+  },
+  "movieParams": {
+    "provider": "replicate",
+    "model": "kwaivgi/kling-v3-video",
+    "concurrency": 1
+  },
+  "beats": [
+    {
+      "text": "Audio ON (generateAudio: true)",
+      "moviePrompt": "A cat meowing and walking across a wooden floor, indoor scene",
+      "movieParams": {
+        "generateAudio": true
+      },
+      "duration": 3
+    },
+    {
+      "text": "Audio OFF (generateAudio: false)",
+      "moviePrompt": "A cat meowing and walking across a wooden floor, indoor scene",
+      "movieParams": {
+        "generateAudio": false
+      },
+      "duration": 3
+    }
+  ]
+}

package/scripts/test/{gpt.json → test_gpt_image.json}

@@ -21,6 +21,14 @@
     }
   },
   "beats": [
+    {
+      "speaker": "Host",
+      "text": "How are you?",
+      "imagePrompt": "A witch in Harajuku",
+      "imageParams": {
+        "model": "gpt-image-2"
+      }
+    },
     {
       "speaker": "Host",
       "text": "How are you?",

package/scripts/test/test_seedance.json

@@ -0,0 +1,61 @@
+{
+  "$mulmocast": { "version": "1.1" },
+  "movieParams": {
+    "provider": "replicate",
+    "model": "bytedance/seedance-2.0"
+  },
+  "audioParams": {
+    "bgmVolume": 0
+  },
+  "captionParams": {
+    "lang": "en"
+  },
+  "lang": "en",
+  "beats": [
+    {
+      "text": "Comparing the bytedance seedance series with a fast-cut dance prompt",
+      "image": {
+        "type": "textSlide",
+        "slide": {
+          "title": "PROMPT: high-energy dance montage, quick cuts synced to the beat, neon-lit urban street"
+        }
+      }
+    },
+    {
+      "id": "seedance-1-lite",
+      "text": "bytedance/seedance-1-lite",
+      "duration": 5,
+      "moviePrompt": "A high-energy 5-second dance video, photorealistic, vibrant lighting. Fast-paced montage with quick cuts synchronized to the beat: energetic dancer performing sharp hip-hop moves in an urban street at night, neon lights reflecting on wet pavement, quick close-ups on footwork and hand gestures, dynamic full-body shots with camera orbiting smoothly, realistic body physics and fabric movement, original choreography, no copyrighted elements.",
+      "movieParams": {
+        "model": "bytedance/seedance-1-lite"
+      }
+    },
+    {
+      "id": "seedance-1-pro",
+      "text": "bytedance/seedance-1-pro",
+      "duration": 5,
+      "moviePrompt": "A high-energy 5-second dance video, photorealistic, vibrant lighting. Fast-paced montage with quick cuts synchronized to the beat: energetic dancer performing sharp hip-hop moves in an urban street at night, neon lights reflecting on wet pavement, quick close-ups on footwork and hand gestures, dynamic full-body shots with camera orbiting smoothly, realistic body physics and fabric movement, original choreography, no copyrighted elements.",
+      "movieParams": {
+        "model": "bytedance/seedance-1-pro"
+      }
+    },
+    {
+      "id": "seedance-2.0",
+      "text": "bytedance/seedance-2.0",
+      "duration": 10,
+      "moviePrompt": "A high-energy 10-second dance video, photorealistic, vibrant lighting. Fast-paced montage with quick cuts synchronized to the beat: energetic dancer performing sharp hip-hop moves in an urban street at night, neon lights reflecting on wet pavement, quick close-ups on footwork and hand gestures, dynamic full-body shots with camera orbiting smoothly, realistic body physics and fabric movement, original choreography, no copyrighted elements.",
+      "movieParams": {
+        "model": "bytedance/seedance-2.0"
+      }
+    },
+    {
+      "id": "seedance-2.0-fast",
+      "text": "bytedance/seedance-2.0-fast",
+      "duration": 8,
+      "moviePrompt": "A high-energy 8-second dance video, photorealistic, vibrant lighting. Fast-paced montage with quick cuts synchronized to the beat: energetic dancer performing sharp hip-hop moves in an urban street at night, neon lights reflecting on wet pavement, quick close-ups on footwork and hand gestures, dynamic full-body shots with camera orbiting smoothly, realistic body physics and fabric movement, original choreography, no copyrighted elements.",
+      "movieParams": {
+        "model": "bytedance/seedance-2.0-fast"
+      }
+    }
+  ]
+}

package/scripts/test/test_seedance2_only.json

@@ -0,0 +1,43 @@
+{
+  "$mulmocast": { "version": "1.1" },
+  "movieParams": {
+    "provider": "replicate",
+    "model": "bytedance/seedance-2.0"
+  },
+  "audioParams": {
+    "bgmVolume": 0
+  },
+  "captionParams": {
+    "lang": "en"
+  },
+  "lang": "en",
+  "beats": [
+    {
+      "text": "Comparing the bytedance seedance series with a fast-cut dance prompt",
+      "image": {
+        "type": "textSlide",
+        "slide": {
+          "title": "PROMPT: high-energy dance montage, quick cuts synced to the beat, neon-lit urban street"
+        }
+      }
+    },
+    {
+      "id": "seedance-2.0",
+      "text": "bytedance/seedance-2.0",
+      "duration": 5,
+      "moviePrompt": "A high-energy 5-second dance video, photorealistic, vibrant lighting. Fast-paced montage with quick cuts synchronized to the beat: energetic dancer performing sharp hip-hop moves in an urban street at night, neon lights reflecting on wet pavement, quick close-ups on footwork and hand gestures, dynamic full-body shots with camera orbiting smoothly, realistic body physics and fabric movement, original choreography, no copyrighted elements.",
+      "movieParams": {
+        "model": "bytedance/seedance-2.0"
+      }
+    },
+    {
+      "id": "seedance-2.0-fast",
+      "text": "bytedance/seedance-2.0-fast",
+      "duration": 5,
+      "moviePrompt": "A high-energy 5-second dance video, photorealistic, vibrant lighting. Fast-paced montage with quick cuts synchronized to the beat: energetic dancer performing sharp hip-hop moves in an urban street at night, neon lights reflecting on wet pavement, quick close-ups on footwork and hand gestures, dynamic full-body shots with camera orbiting smoothly, realistic body physics and fabric movement, original choreography, no copyrighted elements.",
+      "movieParams": {
+        "model": "bytedance/seedance-2.0-fast"
+      }
+    }
+  ]
+}