mulmocast 2.6.6 → 2.6.8

package/lib/actions/audio.js

@@ -1,12 +1,11 @@
 import dotenv from "dotenv";
-import { GraphAI, TaskManager, GraphAILogger } from "graphai";
+import { GraphAI, GraphAILogger } from "graphai";
 import * as agents from "@graphai/vanilla";
 import { fileWriteAgent } from "@graphai/vanilla_node_agents";
 import { ttsOpenaiAgent, ttsGoogleAgent, ttsGeminiAgent, ttsElevenlabsAgent, ttsKotodamaAgent, addBGMAgent, combineAudioFilesAgent, mediaMockAgent, } from "../agents/index.js";
-import { text2SpeechProviderSchema } from "../types/index.js";
-import { fileCacheAgentFilter } from "../utils/filters.js";
+import { audioGraphOption } from "./graph_option.js";
 import { getAudioArtifactFilePath, getAudioFilePath, getGroupedAudioFilePath, getOutputStudioFilePath, resolveDirPath, defaultBGMPath, mkdir, writingMessage, } from "../utils/file.js";
-import { localizedText, settings2GraphAIConfig } from "../utils/utils.js";
+import { localizedText } from "../utils/utils.js";
 import { text2hash } from "../utils/utils_node.js";
 import { provider2TTSAgent } from "../types/provider2agent.js";
 import { invalidAudioSourceError } from "../utils/error_cause.js";
@@ -210,21 +209,6 @@ export const audio_graph_data = {
         },
     },
 };
-const agentFilters = [
-    {
-        name: "fileCacheAgentFilter",
-        agent: fileCacheAgentFilter,
-        nodeIds: ["tts"],
-    },
-];
-const getConcurrency = (context) => {
-    // Check if any speaker uses elevenlabs or kotodama (providers that require concurrency = 1)
-    const hasLimitedConcurrencyProvider = Object.values(context.presentationStyle.speechParams.speakers).some((speaker) => {
-        const provider = text2SpeechProviderSchema.parse(speaker.provider);
-        return provider2TTSAgent[provider].hasLimitedConcurrency;
-    });
-    return hasLimitedConcurrencyProvider ? 1 : 8;
-};
 const audioAgents = {
     ...vanillaAgents,
     fileWriteAgent,
@@ -246,9 +230,8 @@ export const generateBeatAudio = async (index, context, args) => {
         const audioSegmentDirPath = context.fileDirs.grouped ? audioDirPath : resolveDirPath(audioDirPath, fileName);
         mkdir(outDirPath);
         mkdir(audioSegmentDirPath);
-        const config = settings2GraphAIConfig(settings);
-        const taskManager = new TaskManager(getConcurrency(context));
-        const graph = new GraphAI(langs ? graph_tts_map : graph_tts, audioAgents, { agentFilters, taskManager, config });
+        const graph = new GraphAI(langs ? graph_tts_map : graph_tts, audioAgents, await audioGraphOption(context, settings));
+        callbacks?.forEach((callback) => graph.registerCallback(callback));
         graph.injectValue("__mapIndex", index);
         graph.injectValue("beat", context.studio.script.beats[index]);
         graph.injectValue("studioBeat", context.studio.beats[index]);
@@ -260,11 +243,6 @@ export const generateBeatAudio = async (index, context, args) => {
         else {
             graph.injectValue("lang", context.lang);
         }
-        if (callbacks) {
-            callbacks.forEach((callback) => {
-                graph.registerCallback(callback);
-            });
-        }
         await graph.run();
     }
     catch (error) {
@@ -288,19 +266,13 @@ export const audio = async (context, args) => {
         const outputStudioFilePath = getOutputStudioFilePath(outDirPath, fileName);
         mkdir(outDirPath);
         mkdir(audioSegmentDirPath);
-        const config = settings2GraphAIConfig(settings, process.env);
-        const taskManager = new TaskManager(getConcurrency(context));
-        const graph = new GraphAI(audio_graph_data, audioAgents, { agentFilters, taskManager, config });
+        const graph = new GraphAI(audio_graph_data, audioAgents, await audioGraphOption(context, settings));
+        callbacks?.forEach((callback) => graph.registerCallback(callback));
         graph.injectValue("context", context);
         graph.injectValue("audioArtifactFilePath", audioArtifactFilePath);
         graph.injectValue("audioCombinedFilePath", audioCombinedFilePath);
         graph.injectValue("outputStudioFilePath", outputStudioFilePath);
         graph.injectValue("musicFile", MulmoMediaSourceMethods.resolve(context.presentationStyle.audioParams.bgm, context) ?? process.env.PATH_BGM ?? defaultBGMPath());
-        if (callbacks) {
-            callbacks.forEach((callback) => {
-                graph.registerCallback(callback);
-            });
-        }
         const result = await graph.run();
         writingMessage(audioCombinedFilePath);
         MulmoStudioContextMethods.setSessionState(context, "audio", false, true);

package/lib/actions/graph_option.d.ts

@@ -1,3 +1,4 @@
 import type { GraphOptions } from "graphai";
 import { MulmoStudioContext } from "../types/index.js";
-export declare const graphOption: (context: MulmoStudioContext, settings?: Record<string, string>) => Promise<GraphOptions>;
+export declare const imageGraphOption: (context: MulmoStudioContext, settings?: Record<string, string>) => Promise<GraphOptions>;
+export declare const audioGraphOption: (context: MulmoStudioContext, settings?: Record<string, string>) => Promise<GraphOptions>;

package/lib/actions/graph_option.js

@@ -2,17 +2,18 @@ import { TaskManager } from "graphai";
 import { MulmoPresentationStyleMethods } from "../methods/index.js";
 import { fileCacheAgentFilter } from "../utils/filters.js";
 import { settings2GraphAIConfig } from "../utils/utils.js";
-export const graphOption = async (context, settings) => {
-    const options = {
-        agentFilters: [
-            {
-                name: "fileCacheAgentFilter",
-                agent: fileCacheAgentFilter,
-                nodeIds: ["imageGenerator", "movieGenerator", "htmlImageAgent", "soundEffectGenerator", "lipSyncGenerator", "AudioTrimmer"],
-            },
-        ],
-        taskManager: new TaskManager(MulmoPresentationStyleMethods.getConcurrency(context.presentationStyle)),
-        config: settings2GraphAIConfig(settings, process.env),
-    };
-    return options;
-};
+const createGraphOption = (concurrency, cacheNodeIds, settings) => ({
+    agentFilters: [
+        {
+            name: "fileCacheAgentFilter",
+            agent: fileCacheAgentFilter,
+            nodeIds: cacheNodeIds,
+        },
+    ],
+    taskManager: new TaskManager(concurrency),
+    config: settings2GraphAIConfig(settings, process.env),
+});
+const IMAGE_CACHE_NODE_IDS = ["imageGenerator", "movieGenerator", "htmlImageAgent", "soundEffectGenerator", "lipSyncGenerator", "AudioTrimmer"];
+const AUDIO_CACHE_NODE_IDS = ["tts"];
+export const imageGraphOption = async (context, settings) => createGraphOption(MulmoPresentationStyleMethods.getImageConcurrency(context.presentationStyle), IMAGE_CACHE_NODE_IDS, settings);
+export const audioGraphOption = async (context, settings) => createGraphOption(MulmoPresentationStyleMethods.getAudioConcurrency(context.presentationStyle), AUDIO_CACHE_NODE_IDS, settings);

package/lib/actions/image_references.js

@@ -1,6 +1,6 @@
 import { GraphAI, GraphAILogger } from "graphai";
 import { getReferenceImagePath } from "../utils/file.js";
-import { graphOption } from "./graph_option.js";
+import { imageGraphOption } from "./graph_option.js";
 import { MulmoPresentationStyleMethods, MulmoMediaSourceMethods } from "../methods/index.js";
 import { imageOpenaiAgent, mediaMockAgent, imageGenAIAgent, imageReplicateAgent, movieGenAIAgent, movieReplicateAgent } from "../agents/index.js";
 import { agentGenerationError, imageReferenceAction, imageFileTarget, movieFileTarget } from "../utils/error_cause.js";
@@ -41,7 +41,7 @@ export const generateReferenceImage = async (inputs) => {
         },
     };
     try {
-        const options = await graphOption(context);
+        const options = await imageGraphOption(context);
         const graph = new GraphAI(image_graph_data, { imageGenAIAgent, imageOpenaiAgent, mediaMockAgent, imageReplicateAgent }, options);
         await graph.run();
         return imagePath;
@@ -121,12 +121,13 @@ const generateReferenceMovie = async (inputs) => {
                 params: {
                     model: movieAgentInfo.movieParams.model,
                     canvasSize: context.presentationStyle.canvasSize,
+                    generateAudio: movieAgentInfo.movieParams.generateAudio,
                 },
             },
         },
     };
     try {
-        const options = await graphOption(context);
+        const options = await imageGraphOption(context);
         const graph = new GraphAI(movie_graph_data, { movieGenAIAgent, movieReplicateAgent, mediaMockAgent }, options);
         await graph.run();
         return moviePath;

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: {};
@@ -463,7 +464,7 @@ export declare const beat_graph_data: {
     };
 };
 export declare const images_graph_data: GraphData;
-export { graphOption } from "./graph_option.js";
+export { imageGraphOption } from "./graph_option.js";
 type ImageOptions = {
     imageAgents: Record<string, unknown>;
 };

package/lib/actions/images.js

@@ -14,7 +14,7 @@ import { audioCheckerError } from "../utils/error_cause.js";
 import { extractImageFromMovie, ffmpegGetMediaDuration, trimMusic } from "../utils/ffmpeg_utils.js";
 import { getMediaRefs, resolveBeatLocalRefs } from "./image_references.js";
 import { imagePreprocessAgent, imagePluginAgent, htmlImageGeneratorAgent } from "./image_agents.js";
-import { graphOption } from "./graph_option.js";
+import { imageGraphOption } from "./graph_option.js";
 const vanillaAgents = vanilla.default ?? vanilla;
 const imageAgents = {
     imageGenAIAgent,
@@ -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: {},
@@ -431,8 +432,7 @@ export const images_graph_data = {
         },
     },
 };
-// graphOption moved to graph_option.ts to break circular dependency with image_references.ts
-export { graphOption } from "./graph_option.js";
+export { imageGraphOption } from "./graph_option.js";
 const prepareGenerateImages = async (context) => {
     const fileName = MulmoStudioContextMethods.getFileName(context);
     const imageProjectDirPath = MulmoStudioContextMethods.getImageProjectDirPath(context);
@@ -459,7 +459,7 @@ const generateImages = async (context, args) => {
         ...defaultAgents,
         ...optionImageAgents,
     };
-    const graph = new GraphAI(images_graph_data, graphaiAgent, await graphOption(context, settings));
+    const graph = new GraphAI(images_graph_data, graphaiAgent, await imageGraphOption(context, settings));
     Object.keys(injections).forEach((key) => {
         graph.injectValue(key, injections[key]);
     });
@@ -499,7 +499,7 @@ export const generateBeatImage = async (inputs) => {
     try {
         const { index, context, args } = inputs;
         const { settings, callbacks, forceMovie, forceImage, forceLipSync, forceSoundEffect, withBackup } = args ?? {};
-        const options = await graphOption(context, settings);
+        const options = await imageGraphOption(context, settings);
         const injections = await prepareGenerateImages(context);
         const graph = new GraphAI(beat_graph_data, defaultAgents, options);
         Object.keys(injections).forEach((key) => {

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/methods/mulmo_presentation_style.d.ts

@@ -179,6 +179,8 @@ export declare const MulmoPresentationStyleMethods: {
                 imageName: string;
                 referenceType: "ASSET" | "STYLE";
             }[] | undefined;
+            concurrency?: number | undefined;
+            generateAudio?: boolean | undefined;
             speed?: number | undefined;
         };
         keyName: string;
@@ -204,7 +206,10 @@ export declare const MulmoPresentationStyleMethods: {
             image?: string;
         }>;
     };
-    getConcurrency(presentationStyle: MulmoPresentationStyle): 4 | 16;
+    /** Concurrency for image/movie generation graph (uses min of imageParams/movieParams) */
+    getImageConcurrency(presentationStyle: MulmoPresentationStyle): number;
+    /** Concurrency for audio/TTS generation graph */
+    getAudioConcurrency(presentationStyle: MulmoPresentationStyle): number;
     getHtmlImageAgentInfo(presentationStyle: MulmoPresentationStyle): Text2HtmlAgentInfo;
     getImageType(_: MulmoPresentationStyle, beat: MulmoBeat): BeatMediaType;
 };

package/lib/methods/mulmo_presentation_style.js

@@ -6,10 +6,10 @@
 import { isNull } from "graphai";
 import { userAssert } from "../utils/utils.js";
 import { text2ImageProviderSchema, text2HtmlImageProviderSchema, text2MovieProviderSchema, text2SpeechProviderSchema, mulmoCanvasDimensionSchema, mulmoTransitionSchema, } from "../types/schema.js";
-import { provider2ImageAgent, provider2MovieAgent, provider2LLMAgent, provider2SoundEffectAgent, provider2LipSyncAgent, defaultProviders, } from "../types/provider2agent.js";
+import { provider2ImageAgent, provider2MovieAgent, provider2LLMAgent, provider2TTSAgent, provider2SoundEffectAgent, provider2LipSyncAgent, defaultProviders, } from "../types/provider2agent.js";
 const defaultTextSlideStyles = [
     '*,*::before,*::after{box-sizing:border-box}body,h1,h2,h3,h4,p,figure,blockquote,dl,dd{margin:0}ul[role="list"],ol[role="list"]{list-style:none}html:focus-within{scroll-behavior:smooth}body{min-height:100vh;text-rendering:optimizeSpeed;line-height:1.5}a:not([class]){text-decoration-skip-ink:auto}img,picture{max-width:100%;display:block}input,button,textarea,select{font:inherit}@media(prefers-reduced-motion:reduce){html:focus-within{scroll-behavior:auto}*,*::before,*::after{animation-duration:.01ms !important;animation-iteration-count:1 !important;transition-duration:.01ms !important;scroll-behavior:auto !important}}',
-    "body { margin: 60px; margin-top: 40px; color:#333; font-size: 30px; font-family: Arial, sans-serif; box-sizing: border-box; height: 100vh }",
+    "body { margin: 60px; margin-top: 40px; color:#333; background-color:#fff; font-size: 30px; font-family: Arial, sans-serif; box-sizing: border-box; height: 100vh }",
     "h1 { font-size: 56px; margin-bottom: 20px; text-align: center }",
     "h2 { font-size: 48px; text-align: center }",
     "h3 { font-size: 36px }",
@@ -119,7 +119,17 @@ export const MulmoPresentationStyleMethods = {
         const agentInfo = provider2LipSyncAgent[lipSyncProvider];
         return agentInfo;
     },
-    getConcurrency(presentationStyle) {
+    /** Concurrency for image/movie generation graph (uses min of imageParams/movieParams) */
+    getImageConcurrency(presentationStyle) {
+        const imageConcurrency = presentationStyle.imageParams?.concurrency;
+        const movieConcurrency = presentationStyle.movieParams?.concurrency;
+        // User-specified concurrency takes precedence.
+        // Use the smaller of imageParams/movieParams since they share the same graph.
+        if (imageConcurrency !== undefined || movieConcurrency !== undefined) {
+            const values = [imageConcurrency, movieConcurrency].filter((v) => v !== undefined);
+            return Math.min(...values);
+        }
+        // Fallback: provider-based auto-detection
         const imageAgentInfo = MulmoPresentationStyleMethods.getImageAgentInfo(presentationStyle);
         if (imageAgentInfo.imageParams.provider === "openai") {
             // NOTE: Here are the rate limits of OpenAI's text2image API (1token = 32x32 patch).
@@ -131,6 +141,20 @@ export const MulmoPresentationStyleMethods = {
         }
         return 4;
     },
+    /** Concurrency for audio/TTS generation graph */
+    getAudioConcurrency(presentationStyle) {
+        // User-specified concurrency takes precedence
+        const userConcurrency = presentationStyle.audioParams?.concurrency;
+        if (userConcurrency !== undefined) {
+            return userConcurrency;
+        }
+        // Fallback: provider-based auto-detection
+        const hasLimitedConcurrencyProvider = Object.values(presentationStyle.speechParams.speakers).some((speaker) => {
+            const provider = text2SpeechProviderSchema.parse(speaker.provider);
+            return provider2TTSAgent[provider].hasLimitedConcurrency;
+        });
+        return hasLimitedConcurrencyProvider ? 1 : 8;
+    },
     getHtmlImageAgentInfo(presentationStyle) {
         const provider = text2HtmlImageProviderSchema.parse(presentationStyle.htmlImageParams?.provider);
         const defaultConfig = provider2LLMAgent[provider];

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

@@ -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

@@ -3299,6 +3299,7 @@ export declare const mulmoImageParamsSchema: z.ZodObject<{
         }>>;
         opacity: z.ZodOptional<z.ZodNumber>;
     }, z.core.$strip>]>>>;
+    concurrency: z.ZodOptional<z.ZodNumber>;
 }, z.core.$strict>;
 export declare const textSlideParamsSchema: z.ZodObject<{
     cssStyles: z.ZodUnion<readonly [z.ZodString, z.ZodArray<z.ZodString>]>;
@@ -3398,6 +3399,7 @@ export declare const audioParamsSchema: z.ZodObject<{
     ducking: z.ZodOptional<z.ZodObject<{
         ratio: z.ZodOptional<z.ZodNumber>;
     }, z.core.$strip>>;
+    concurrency: z.ZodOptional<z.ZodNumber>;
 }, z.core.$strict>;
 export declare const htmlPromptParamsSchema: z.ZodObject<{
     systemPrompt: z.ZodDefault<z.ZodOptional<z.ZodString>>;
@@ -3688,6 +3690,8 @@ export declare const mulmoMovieParamsSchema: z.ZodObject<{
             STYLE: "STYLE";
         }>;
     }, 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>;
@@ -6581,6 +6585,8 @@ export declare const mulmoBeatSchema: z.ZodObject<{
                 STYLE: "STYLE";
             }>;
         }, 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<{
@@ -6829,6 +6835,7 @@ export declare const mulmoPresentationStyleSchema: z.ZodObject<{
             }>>;
             opacity: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strip>]>>>;
+        concurrency: z.ZodOptional<z.ZodNumber>;
     }, z.core.$strict>>>;
     movieParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
         provider: z.ZodOptional<z.ZodEnum<{
@@ -7068,6 +7075,8 @@ export declare const mulmoPresentationStyleSchema: z.ZodObject<{
                 STYLE: "STYLE";
             }>;
         }, 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<{
@@ -7191,6 +7200,7 @@ export declare const mulmoPresentationStyleSchema: z.ZodObject<{
         ducking: z.ZodOptional<z.ZodObject<{
             ratio: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strip>>;
+        concurrency: z.ZodOptional<z.ZodNumber>;
     }, z.core.$strict>>;
 }, z.core.$strip>;
 export declare const mulmoReferenceSchema: z.ZodObject<{
@@ -7330,6 +7340,7 @@ export declare const mulmoScriptSchema: z.ZodObject<{
             }>>;
             opacity: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strip>]>>>;
+        concurrency: z.ZodOptional<z.ZodNumber>;
     }, z.core.$strict>>>;
     movieParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
         provider: z.ZodOptional<z.ZodEnum<{
@@ -7569,6 +7580,8 @@ export declare const mulmoScriptSchema: z.ZodObject<{
                 STYLE: "STYLE";
             }>;
         }, z.core.$strip>>>;
+        concurrency: z.ZodOptional<z.ZodNumber>;
+        generateAudio: z.ZodOptional<z.ZodBoolean>;
     }, z.core.$strip>>>;
     soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
         provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{
@@ -7692,6 +7705,7 @@ export declare const mulmoScriptSchema: z.ZodObject<{
         ducking: z.ZodOptional<z.ZodObject<{
             ratio: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strip>>;
+        concurrency: z.ZodOptional<z.ZodNumber>;
     }, z.core.$strict>>;
     title: z.ZodOptional<z.ZodString>;
     description: z.ZodOptional<z.ZodString>;
@@ -10600,6 +10614,8 @@ export declare const mulmoScriptSchema: z.ZodObject<{
                     STYLE: "STYLE";
                 }>;
             }, z.core.$strip>>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
+            generateAudio: z.ZodOptional<z.ZodBoolean>;
             speed: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strip>>;
         soundEffectParams: z.ZodOptional<z.ZodObject<{
@@ -10923,6 +10939,7 @@ export declare const mulmoStudioSchema: z.ZodObject<{
                 }>>;
                 opacity: z.ZodOptional<z.ZodNumber>;
             }, z.core.$strip>]>>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strict>>>;
         movieParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodEnum<{
@@ -11162,6 +11179,8 @@ export declare const mulmoStudioSchema: z.ZodObject<{
                     STYLE: "STYLE";
                 }>;
             }, z.core.$strip>>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
+            generateAudio: z.ZodOptional<z.ZodBoolean>;
         }, z.core.$strip>>>;
         soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{
@@ -11285,6 +11304,7 @@ export declare const mulmoStudioSchema: z.ZodObject<{
             ducking: z.ZodOptional<z.ZodObject<{
                 ratio: z.ZodOptional<z.ZodNumber>;
             }, z.core.$strip>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strict>>;
         title: z.ZodOptional<z.ZodString>;
         description: z.ZodOptional<z.ZodString>;
@@ -14193,6 +14213,8 @@ export declare const mulmoStudioSchema: z.ZodObject<{
                         STYLE: "STYLE";
                     }>;
                 }, z.core.$strip>>>;
+                concurrency: z.ZodOptional<z.ZodNumber>;
+                generateAudio: z.ZodOptional<z.ZodBoolean>;
                 speed: z.ZodOptional<z.ZodNumber>;
             }, z.core.$strip>>;
             soundEffectParams: z.ZodOptional<z.ZodObject<{
@@ -14452,6 +14474,7 @@ export declare const mulmoPromptTemplateSchema: z.ZodObject<{
                 }>>;
                 opacity: z.ZodOptional<z.ZodNumber>;
             }, z.core.$strip>]>>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strict>>>;
         movieParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodEnum<{
@@ -14691,6 +14714,8 @@ export declare const mulmoPromptTemplateSchema: z.ZodObject<{
                     STYLE: "STYLE";
                 }>;
             }, z.core.$strip>>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
+            generateAudio: z.ZodOptional<z.ZodBoolean>;
         }, z.core.$strip>>>;
         soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{
@@ -14814,6 +14839,7 @@ export declare const mulmoPromptTemplateSchema: z.ZodObject<{
             ducking: z.ZodOptional<z.ZodObject<{
                 ratio: z.ZodOptional<z.ZodNumber>;
             }, z.core.$strip>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strict>>;
     }, z.core.$strip>>;
 }, z.core.$strict>;
@@ -14947,6 +14973,7 @@ export declare const mulmoPromptTemplateFileSchema: z.ZodObject<{
                 }>>;
                 opacity: z.ZodOptional<z.ZodNumber>;
             }, z.core.$strip>]>>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strict>>>;
         movieParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodEnum<{
@@ -15186,6 +15213,8 @@ export declare const mulmoPromptTemplateFileSchema: z.ZodObject<{
                     STYLE: "STYLE";
                 }>;
             }, z.core.$strip>>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
+            generateAudio: z.ZodOptional<z.ZodBoolean>;
         }, z.core.$strip>>>;
         soundEffectParams: z.ZodDefault<z.ZodOptional<z.ZodObject<{
             provider: z.ZodOptional<z.ZodDefault<z.ZodEnum<{
@@ -15309,6 +15338,7 @@ export declare const mulmoPromptTemplateFileSchema: z.ZodObject<{
             ducking: z.ZodOptional<z.ZodObject<{
                 ratio: z.ZodOptional<z.ZodNumber>;
             }, z.core.$strip>>;
+            concurrency: z.ZodOptional<z.ZodNumber>;
         }, z.core.$strict>>;
     }, z.core.$strip>>;
     filename: z.ZodString;

package/lib/types/schema.js

@@ -393,6 +393,7 @@ export const mulmoImageParamsSchema = mulmoBeatImageParamsSchema
     .extend({
     images: mulmoImageParamsImagesSchema.optional(),
     backgroundImage: backgroundImageSchema,
+    concurrency: z.number().int().positive().optional().describe("Max concurrent image generation requests"),
 })
     .strict();
 export const textSlideParamsSchema = z
@@ -436,6 +437,7 @@ export const audioParamsSchema = z
     })
         .optional()
         .describe("Auto-reduce movie audio when TTS is playing"),
+    concurrency: z.number().int().positive().optional().describe("Max concurrent TTS generation requests"),
 })
     .strict();
 export const htmlPromptParamsSchema = z
@@ -496,6 +498,8 @@ export const mulmoMovieParamsSchema = z.object({
         .array(movieReferenceImageSchema)
         .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

@@ -115,6 +115,7 @@ export declare const createStudioData: (_mulmoScript: MulmoScript, fileName: str
                 size?: "contain" | "cover" | "fill" | "auto" | undefined;
                 opacity?: number | undefined;
             } | null | undefined;
+            concurrency?: number | undefined;
         };
         movieParams: {
             provider?: string | undefined;
@@ -280,6 +281,8 @@ export declare const createStudioData: (_mulmoScript: MulmoScript, fileName: str
                 imageName: string;
                 referenceType: "ASSET" | "STYLE";
             }[] | undefined;
+            concurrency?: number | undefined;
+            generateAudio?: boolean | undefined;
         };
         soundEffectParams: {
             provider?: string | undefined;
@@ -308,6 +311,7 @@ export declare const createStudioData: (_mulmoScript: MulmoScript, fileName: str
             ducking?: {
                 ratio?: number | undefined;
             } | undefined;
+            concurrency?: number | undefined;
         };
         lang: string;
         beats: {
@@ -2018,6 +2022,8 @@ export declare const createStudioData: (_mulmoScript: MulmoScript, fileName: str
                     imageName: string;
                     referenceType: "ASSET" | "STYLE";
                 }[] | undefined;
+                concurrency?: number | undefined;
+                generateAudio?: boolean | undefined;
                 speed?: number | undefined;
             } | undefined;
             soundEffectParams?: {
@@ -2347,6 +2353,7 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                     size?: "contain" | "cover" | "fill" | "auto" | undefined;
                     opacity?: number | undefined;
                 } | null | undefined;
+                concurrency?: number | undefined;
             };
             movieParams: {
                 provider?: string | undefined;
@@ -2512,6 +2519,8 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                     imageName: string;
                     referenceType: "ASSET" | "STYLE";
                 }[] | undefined;
+                concurrency?: number | undefined;
+                generateAudio?: boolean | undefined;
             };
             soundEffectParams: {
                 provider?: string | undefined;
@@ -2540,6 +2549,7 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                 ducking?: {
                     ratio?: number | undefined;
                 } | undefined;
+                concurrency?: number | undefined;
             };
             lang: string;
             beats: {
@@ -4250,6 +4260,8 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                         imageName: string;
                         referenceType: "ASSET" | "STYLE";
                     }[] | undefined;
+                    concurrency?: number | undefined;
+                    generateAudio?: boolean | undefined;
                     speed?: number | undefined;
                 } | undefined;
                 soundEffectParams?: {
@@ -4586,6 +4598,7 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                 size?: "contain" | "cover" | "fill" | "auto" | undefined;
                 opacity?: number | undefined;
             } | null | undefined;
+            concurrency?: number | undefined;
         };
         movieParams: {
             provider?: string | undefined;
@@ -4751,6 +4764,8 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
                 imageName: string;
                 referenceType: "ASSET" | "STYLE";
             }[] | undefined;
+            concurrency?: number | undefined;
+            generateAudio?: boolean | undefined;
         };
         soundEffectParams: {
             provider?: string | undefined;
@@ -4779,6 +4794,7 @@ export declare const initializeContextFromFiles: (files: FileObject, raiseError:
             ducking?: {
                 ratio?: number | undefined;
             } | undefined;
+            concurrency?: number | undefined;
         };
         lipSyncParams?: {
             provider?: string | undefined;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mulmocast",
-  "version": "2.6.6",
+  "version": "2.6.8",
   "description": "",
   "type": "module",
   "main": "lib/index.node.js",
@@ -24,8 +24,9 @@
     }
   },
   "resolutions": {
-    "minimatch": "^10.2.4",
-    "yauzl": "^3.2.1"
+    "minimatch": "^10.2.5",
+    "tar": "7.5.13",
+    "yauzl": "^3.3.0"
   },
   "bin": {
     "mulmo": "lib/cli/bin.js",
@@ -88,7 +89,7 @@
   "homepage": "https://github.com/receptron/mulmocast-cli#readme",
   "dependencies": {
     "@google-cloud/text-to-speech": "^6.4.0",
-    "@google/genai": "^1.48.0",
+    "@google/genai": "^1.50.1",
     "@graphai/anthropic_agent": "^2.0.12",
     "@graphai/browserless_agent": "^2.0.2",
     "@graphai/gemini_agent": "^2.0.5",
@@ -98,21 +99,21 @@
     "@graphai/stream_agent_filter": "^2.0.3",
     "@graphai/vanilla": "^2.0.12",
     "@graphai/vanilla_node_agents": "^2.0.4",
-    "@inquirer/input": "^5.0.10",
-    "@inquirer/select": "^5.1.2",
+    "@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",
     "archiver": "^7.0.1",
     "clipboardy": "^5.3.1",
-    "dotenv": "^17.4.0",
+    "dotenv": "^17.4.2",
     "fluent-ffmpeg": "^2.1.3",
     "graphai": "^2.0.16",
-    "jsdom": "^29.0.1",
-    "marked": "^17.0.5",
+    "jsdom": "^29.0.2",
+    "marked": "^18.0.2",
     "mulmocast-vision": "^1.0.9",
     "ora": "^9.3.0",
-    "puppeteer": "^24.40.0",
+    "puppeteer": "^24.41.0",
     "replicate": "^1.4.0",
     "yaml": "^2.8.3",
     "yargs": "^18.0.0",
@@ -126,16 +127,16 @@
     "@types/jsdom": "^28.0.1",
     "@types/yargs": "^17.0.35",
     "cross-env": "^10.1.0",
-    "eslint": "^10.1.0",
+    "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",
-    "globals": "^17.4.0",
-    "prettier": "^3.8.1",
+    "eslint-plugin-sonarjs": "^4.0.3",
+    "globals": "^17.5.0",
+    "prettier": "^3.8.3",
     "tsx": "^4.21.0",
-    "typescript": "6.0.2",
-    "typescript-eslint": "^8.58.0"
+    "typescript": "6.0.3",
+    "typescript-eslint": "^8.58.1"
   },
   "engines": {
     "node": ">=22.0.0"

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/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"
+      }
+    }
+  ]
+}