mulmocast 2.6.9 → 2.6.11

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -60,14 +60,16 @@ export const imageGenAIAgent = async ({ namedInputs, params, config, }) => {
60
60
  const model = params.model ?? provider2ImageAgent["google"].defaultModel;
61
61
  const apiKey = config?.apiKey;
62
62
  const ai = params.vertexai_project
63
- ? new GoogleGenAI({
64
- vertexai: true,
65
- project: params.vertexai_project,
66
- location: params.vertexai_location ?? "us-central1",
67
- })
63
+ ? (() => {
64
+ const location = params.vertexai_location ?? "us-central1";
65
+ if (model === "gemini-3-pro-image-preview" && location !== "global") {
66
+ GraphAILogger.warn(`imageGenAIAgent: model "${model}" on Vertex AI is only available in location "global", but got "${location}". Set imageParams.vertexai_location to "global".`);
67
+ }
68
+ return new GoogleGenAI({ vertexai: true, project: params.vertexai_project, location });
69
+ })()
68
70
  : (() => {
69
71
  if (!apiKey) {
70
- throw new Error("Google GenAI API key is required (GEMINI_API_KEY)", {
72
+ throw new Error("Google GenAI authentication is required. Either set GEMINI_API_KEY (Gemini API) or specify imageParams.vertexai_project (Vertex AI). See docs/vertexai_en.md or docs/vertexai_ja.md.", {
71
73
  cause: apiKeyMissingError("imageGenAIAgent", imageAction, "GEMINI_API_KEY"),
72
74
  });
73
75
  }
@@ -1,5 +1,6 @@
1
1
  import { AgentFunction, AgentFunctionInfo } from "graphai";
2
2
  import type { AgentBufferResult, OpenAIImageAgentParams, OpenAIImageAgentInputs, OpenAIImageAgentConfig } from "../types/agent.js";
3
+ export declare const buildDeprecatedModelMessage: (model: string) => string | null;
3
4
  export declare const imageOpenaiAgent: AgentFunction<OpenAIImageAgentParams, AgentBufferResult, OpenAIImageAgentInputs, OpenAIImageAgentConfig>;
4
5
  declare const imageOpenaiAgentInfo: AgentFunctionInfo;
5
6
  export default imageOpenaiAgentInfo;
@@ -3,8 +3,14 @@ import path from "path";
3
3
  import { GraphAILogger } from "graphai";
4
4
  import { toFile, AuthenticationError, RateLimitError, APIError } from "openai";
5
5
  import { createOpenAIClient } from "../utils/openai_client.js";
6
- import { provider2ImageAgent, gptImages } from "../types/provider2agent.js";
7
- import { apiKeyMissingError, agentGenerationError, openAIAgentGenerationError, agentIncorrectAPIKeyError, agentAPIRateLimitError, agentInvalidResponseError, imageAction, imageFileTarget, } from "../utils/error_cause.js";
6
+ import { provider2ImageAgent, gptImages, deprecatedOpenAIImageModelHints } from "../types/provider2agent.js";
7
+ import { apiKeyMissingError, agentGenerationError, openAIAgentGenerationError, agentIncorrectAPIKeyError, agentAPIRateLimitError, agentInvalidResponseError, imageAction, imageFileTarget, unsupportedModelTarget, } from "../utils/error_cause.js";
8
+ const isDeprecatedOpenAIImageModel = (model) => model in deprecatedOpenAIImageModelHints;
9
+ export const buildDeprecatedModelMessage = (model) => {
10
+ if (!isDeprecatedOpenAIImageModel(model))
11
+ return null;
12
+ return `OpenAI image model "${model}" is no longer available. ${deprecatedOpenAIImageModelHints[model]}`;
13
+ };
8
14
  // https://platform.openai.com/docs/guides/image-generation
9
15
  export const imageOpenaiAgent = async ({ namedInputs, params, config, }) => {
10
16
  const { prompt, referenceImages } = namedInputs;
@@ -16,6 +22,12 @@ export const imageOpenaiAgent = async ({ namedInputs, params, config, }) => {
16
22
  });
17
23
  }
18
24
  const model = params.model ?? provider2ImageAgent["openai"].defaultModel;
25
+ const deprecatedMessage = buildDeprecatedModelMessage(model);
26
+ if (deprecatedMessage) {
27
+ throw new Error(deprecatedMessage, {
28
+ cause: agentGenerationError("imageOpenaiAgent", imageAction, unsupportedModelTarget),
29
+ });
30
+ }
19
31
  const openai = createOpenAIClient({ apiKey, baseURL, apiVersion });
20
32
  const size = (() => {
21
33
  if (gptImages.includes(model)) {
@@ -114,7 +126,7 @@ export const imageOpenaiAgent = async ({ namedInputs, params, config, }) => {
114
126
  }
115
127
  return { buffer: Buffer.from(image_base64, "base64") };
116
128
  }
117
- // For dall-e-3
129
+ // URL response handling (legacy OpenAI image API response format)
118
130
  const res = await fetch(url);
119
131
  if (!res.ok) {
120
132
  throw new Error(`Failed to fetch ${url}: ${res.status} ${res.statusText}`, {
@@ -4,6 +4,22 @@ import Replicate from "replicate";
4
4
  import { getAspectRatio } from "./movie_replicate_agent.js";
5
5
  import { apiKeyMissingError, agentIncorrectAPIKeyError, agentGenerationError, agentInvalidResponseError, imageAction, imageFileTarget, hasCause, } from "../utils/error_cause.js";
6
6
  import { provider2ImageAgent } from "../types/provider2agent.js";
7
+ // Replicate image models return one of: FileOutput (object with url() method),
8
+ // Array<FileOutput>, string URL, or { url: string }. Normalize to URL string/URL.
9
+ const extractImageUrl = (output) => {
10
+ if (typeof output === "string")
11
+ return output;
12
+ if (Array.isArray(output))
13
+ return output.length > 0 ? extractImageUrl(output[0]) : undefined;
14
+ if (output && typeof output === "object" && "url" in output) {
15
+ const url = output.url;
16
+ if (typeof url === "function")
17
+ return url();
18
+ if (typeof url === "string")
19
+ return url;
20
+ }
21
+ return undefined;
22
+ };
7
23
  export const imageReplicateAgent = async ({ namedInputs, params, config, }) => {
8
24
  const { prompt, referenceImages } = namedInputs;
9
25
  const { canvasSize } = params;
@@ -29,22 +45,21 @@ export const imageReplicateAgent = async ({ namedInputs, params, config, }) => {
29
45
  }
30
46
  try {
31
47
  const output = await replicate.run(model, { input });
32
- // Download the generated video
33
- if (output && Array.isArray(output) && output.length > 0 && typeof output[0] === "object" && "url" in output[0]) {
34
- const imageUrl = output[0].url();
35
- const imageResponse = await fetch(imageUrl);
36
- if (!imageResponse.ok) {
37
- throw new Error(`Error downloading image: ${imageResponse.status} - ${imageResponse.statusText}`, {
38
- cause: agentGenerationError("imageReplicateAgent", imageAction, imageFileTarget),
39
- });
40
- }
41
- const arrayBuffer = await imageResponse.arrayBuffer();
42
- const buffer = Buffer.from(arrayBuffer);
43
- return { buffer };
48
+ const imageUrl = extractImageUrl(output);
49
+ if (!imageUrl) {
50
+ throw new Error("ERROR: generateImage returned undefined", {
51
+ cause: agentInvalidResponseError("imageReplicateAgent", imageAction, imageFileTarget),
52
+ });
44
53
  }
45
- throw new Error("ERROR: generateImage returned undefined", {
46
- cause: agentInvalidResponseError("imageReplicateAgent", imageAction, imageFileTarget),
47
- });
54
+ const imageResponse = await fetch(imageUrl);
55
+ if (!imageResponse.ok) {
56
+ throw new Error(`Error downloading image: ${imageResponse.status} - ${imageResponse.statusText}`, {
57
+ cause: agentGenerationError("imageReplicateAgent", imageAction, imageFileTarget),
58
+ });
59
+ }
60
+ const arrayBuffer = await imageResponse.arrayBuffer();
61
+ const buffer = Buffer.from(arrayBuffer);
62
+ return { buffer };
48
63
  }
49
64
  catch (error) {
50
65
  GraphAILogger.info("Replicate generation error:", error);
@@ -175,7 +175,7 @@ export const movieGenAIAgent = async ({ namedInputs, params, config, }) => {
175
175
  })
176
176
  : (() => {
177
177
  if (!apiKey) {
178
- throw new Error("Google GenAI API key is required (GEMINI_API_KEY)", {
178
+ throw new Error("Google GenAI authentication is required. Either set GEMINI_API_KEY (Gemini API) or specify movieParams.vertexai_project (Vertex AI). See docs/vertexai_en.md or docs/vertexai_ja.md.", {
179
179
  cause: apiKeyMissingError("movieGenAIAgent", imageAction, "GEMINI_API_KEY"),
180
180
  });
181
181
  }
@@ -136,6 +136,11 @@ export const movieReplicateAgent = async ({ namedInputs, params, config, }) => {
136
136
  cause: agentGenerationError("movieReplicateAgent", imageAction, unsupportedModelTarget),
137
137
  });
138
138
  }
139
+ if (provider2MovieAgent.replicate.modelParams[model].start_image_required && !imagePath) {
140
+ throw new Error(`Model ${model} requires a start image (image-to-video only)`, {
141
+ cause: agentGenerationError("movieReplicateAgent", imageAction, unsupportedModelTarget),
142
+ });
143
+ }
139
144
  const duration = getModelDuration("replicate", model, params.duration);
140
145
  if (duration === undefined || !provider2MovieAgent.replicate.modelParams[model].durations.includes(duration)) {
141
146
  throw new Error(`Duration ${duration} is not supported for model ${model}. Supported durations: ${provider2MovieAgent.replicate.modelParams[model].durations.join(", ")}`, {
@@ -84,7 +84,7 @@ export const MulmoPresentationStyleMethods = {
84
84
  const imageParams = { ...presentationStyle.imageParams, ...beat?.imageParams };
85
85
  const provider = MulmoPresentationStyleMethods.getText2ImageProvider(imageParams?.provider) ?? defaultProviders.text2image;
86
86
  const agentInfo = provider2ImageAgent[provider];
87
- // The default text2image model is gpt-image-1 from OpenAI, and to use it you must have an OpenAI account and have verified your identity. If this is not possible, please specify dall-e-3 as the model.
87
+ // The default text2image model is gpt-image-1 from OpenAI.
88
88
  const defaultImageParams = {
89
89
  provider,
90
90
  model: agentInfo.defaultModel,
@@ -133,8 +133,7 @@ export const MulmoPresentationStyleMethods = {
133
133
  const imageAgentInfo = MulmoPresentationStyleMethods.getImageAgentInfo(presentationStyle);
134
134
  if (imageAgentInfo.imageParams.provider === "openai") {
135
135
  // NOTE: Here are the rate limits of OpenAI's text2image API (1token = 32x32 patch).
136
- // dall-e-3: 7,500 RPM、15 images per minute (4 images for max resolution)
137
- // gpt-image-1:3,000,000 TPM、150 images per minute
136
+ // gpt-image-1: 3,000,000 TPM, 150 images per minute
138
137
  if (imageAgentInfo.imageParams.model === provider2ImageAgent.openai.defaultModel) {
139
138
  return 16;
140
139
  }
@@ -42,6 +42,11 @@ export declare const provider2TTSAgent: {
42
42
  };
43
43
  };
44
44
  export declare const gptImages: string[];
45
+ export declare const deprecatedOpenAIImageModelHints: {
46
+ readonly "dall-e-2": "Use 'gpt-image-1' or another supported model.";
47
+ readonly "dall-e-3": "Use 'gpt-image-1' or another supported model.";
48
+ };
49
+ export type DeprecatedOpenAIImageModel = keyof typeof deprecatedOpenAIImageModelHints;
45
50
  export declare const provider2ImageAgent: {
46
51
  openai: {
47
52
  agentName: string;
@@ -84,6 +89,7 @@ type MovieAudioSpec = {
84
89
  type ReplicateMovieModelParams = {
85
90
  durations: number[];
86
91
  start_image: string | undefined;
92
+ start_image_required?: boolean;
87
93
  last_image?: string;
88
94
  reference_images_param?: string;
89
95
  audio: MovieAudioSpec;
@@ -45,11 +45,16 @@ export const provider2TTSAgent = {
45
45
  },
46
46
  };
47
47
  export const gptImages = ["gpt-image-2", "gpt-image-1.5", "gpt-image-1", "gpt-image-1-mini"];
48
+ const supportedOpenAIImageReplacementHint = "Use 'gpt-image-1' or another supported model.";
49
+ export const deprecatedOpenAIImageModelHints = {
50
+ "dall-e-2": supportedOpenAIImageReplacementHint,
51
+ "dall-e-3": supportedOpenAIImageReplacementHint,
52
+ };
48
53
  export const provider2ImageAgent = {
49
54
  openai: {
50
55
  agentName: "imageOpenaiAgent",
51
56
  defaultModel: "gpt-image-1",
52
- models: ["dall-e-3", ...gptImages],
57
+ models: [...gptImages],
53
58
  keyName: "OPENAI_API_KEY",
54
59
  baseURLKeyName: "OPENAI_BASE_URL",
55
60
  },
@@ -69,7 +74,23 @@ export const provider2ImageAgent = {
69
74
  replicate: {
70
75
  agentName: "imageReplicateAgent",
71
76
  defaultModel: "bytedance/seedream-4",
72
- models: ["bytedance/seedream-4", "qwen/qwen-image"],
77
+ models: [
78
+ "bytedance/seedream-4",
79
+ "qwen/qwen-image",
80
+ "black-forest-labs/flux-2-pro",
81
+ "black-forest-labs/flux-2-dev",
82
+ "black-forest-labs/flux-1.1-pro",
83
+ "black-forest-labs/flux-1.1-pro-ultra",
84
+ "black-forest-labs/flux-pro",
85
+ "black-forest-labs/flux-dev",
86
+ "black-forest-labs/flux-schnell",
87
+ "ideogram-ai/ideogram-v3-turbo",
88
+ "ideogram-ai/ideogram-v3-balanced",
89
+ "ideogram-ai/ideogram-v3-quality",
90
+ "recraft-ai/recraft-v3",
91
+ "stability-ai/stable-diffusion-3.5-large",
92
+ "luma/photon",
93
+ ],
73
94
  keyName: "REPLICATE_API_TOKEN",
74
95
  },
75
96
  mock: {
@@ -112,6 +133,10 @@ export const provider2MovieAgent = {
112
133
  "runwayml/gen-4.5",
113
134
  "kwaivgi/kling-v3-omni-video",
114
135
  "kwaivgi/kling-v3-video",
136
+ "alibaba/happyhorse-1.0",
137
+ "minimax/hailuo-2.3",
138
+ "minimax/hailuo-2.3-fast",
139
+ "pixverse/pixverse-v5",
115
140
  ],
116
141
  modelParams: {
117
142
  "bytedance/seedance-1-lite": {
@@ -273,6 +298,36 @@ export const provider2MovieAgent = {
273
298
  audio: { mode: AUDIO_MODE_OPTIONAL, param: "generate_audio" },
274
299
  price_per_sec: 0.3,
275
300
  },
301
+ // TODO: price_per_sec for the models below is a coarse approximation.
302
+ // Actual Replicate pricing varies by resolution / duration / quality and
303
+ // cannot be expressed as a single per-second number. Verify each model at
304
+ // https://replicate.com/<owner>/<model> when this field starts being consumed.
305
+ "alibaba/happyhorse-1.0": {
306
+ durations: [3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15],
307
+ start_image: "image",
308
+ audio: { mode: AUDIO_MODE_NEVER },
309
+ price_per_sec: 0.05,
310
+ },
311
+ "minimax/hailuo-2.3": {
312
+ durations: [6, 10],
313
+ start_image: "first_frame_image",
314
+ audio: { mode: AUDIO_MODE_NEVER },
315
+ price_per_sec: 0.1,
316
+ },
317
+ "minimax/hailuo-2.3-fast": {
318
+ durations: [6, 10],
319
+ start_image: "first_frame_image",
320
+ start_image_required: true,
321
+ audio: { mode: AUDIO_MODE_NEVER },
322
+ price_per_sec: 0.06,
323
+ },
324
+ "pixverse/pixverse-v5": {
325
+ durations: [5, 8],
326
+ start_image: "image",
327
+ last_image: "last_frame_image",
328
+ audio: { mode: AUDIO_MODE_NEVER },
329
+ price_per_sec: 0.12,
330
+ },
276
331
  },
277
332
  },
278
333
  google: {
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "mulmocast",
3
- "version": "2.6.9",
3
+ "version": "2.6.11",
4
4
  "description": "",
5
5
  "type": "module",
6
6
  "main": "lib/index.node.js",
@@ -88,7 +88,7 @@
88
88
  },
89
89
  "homepage": "https://github.com/receptron/mulmocast-cli#readme",
90
90
  "dependencies": {
91
- "@google-cloud/text-to-speech": "^6.4.0",
91
+ "@google-cloud/text-to-speech": "^6.4.1",
92
92
  "@google/genai": "^1.50.1",
93
93
  "@graphai/anthropic_agent": "^2.0.12",
94
94
  "@graphai/browserless_agent": "^2.0.2",
@@ -109,15 +109,15 @@
109
109
  "dotenv": "^17.4.2",
110
110
  "fluent-ffmpeg": "^2.1.3",
111
111
  "graphai": "^2.0.16",
112
- "jsdom": "^29.1.0",
113
- "marked": "^18.0.2",
112
+ "jsdom": "^29.1.1",
113
+ "marked": "^18.0.3",
114
114
  "mulmocast-vision": "^1.0.9",
115
115
  "ora": "^9.4.0",
116
- "puppeteer": "^24.42.0",
116
+ "puppeteer": "^24.43.0",
117
117
  "replicate": "^1.4.0",
118
- "yaml": "^2.8.3",
118
+ "yaml": "^2.8.4",
119
119
  "yargs": "^18.0.0",
120
- "zod": "^4.3.6"
120
+ "zod": "^4.4.3"
121
121
  },
122
122
  "devDependencies": {
123
123
  "@eslint/js": "^10.0.1",
@@ -127,16 +127,16 @@
127
127
  "@types/jsdom": "^28.0.1",
128
128
  "@types/yargs": "^17.0.35",
129
129
  "cross-env": "^10.1.0",
130
- "eslint": "^10.2.1",
130
+ "eslint": "^10.3.0",
131
131
  "eslint-config-prettier": "^10.1.8",
132
132
  "eslint-plugin-import": "^2.32.0",
133
133
  "eslint-plugin-prettier": "^5.5.5",
134
134
  "eslint-plugin-sonarjs": "^4.0.3",
135
- "globals": "^17.5.0",
135
+ "globals": "^17.6.0",
136
136
  "prettier": "^3.8.3",
137
137
  "tsx": "^4.21.0",
138
138
  "typescript": "6.0.3",
139
- "typescript-eslint": "^8.59.1"
139
+ "typescript-eslint": "^8.59.2"
140
140
  },
141
141
  "engines": {
142
142
  "node": ">=22.0.0"
@@ -0,0 +1,29 @@
1
+ {
2
+ "$mulmocast": {
3
+ "version": "1.1"
4
+ },
5
+ "canvasSize": {
6
+ "width": 1024,
7
+ "height": 1024
8
+ },
9
+ "imageParams": {
10
+ "provider": "replicate",
11
+ "model": "black-forest-labs/flux-2-pro"
12
+ },
13
+ "speechParams": {
14
+ "speakers": {
15
+ "Presenter": {
16
+ "voiceId": "shimmer",
17
+ "provider": "openai"
18
+ }
19
+ }
20
+ },
21
+ "lang": "en",
22
+ "beats": [
23
+ {
24
+ "speaker": "Presenter",
25
+ "text": "Testing flux-2-pro on Replicate.",
26
+ "imagePrompt": "A cinematic photo of a red panda holding a tiny espresso cup, soft window light, shallow depth of field"
27
+ }
28
+ ]
29
+ }
@@ -21,10 +21,10 @@
21
21
  }
22
22
  },
23
23
  {
24
- "text": "Image with Dall-E 3",
24
+ "text": "Image with OpenAI gpt-image-1-mini",
25
25
  "imagePrompt": "Blue sky, a flock of birds",
26
26
  "imageParams": {
27
- "model": "dall-e-3",
27
+ "model": "gpt-image-1-mini",
28
28
  "style": "anime-style",
29
29
  "provider": "openai"
30
30
  }
@@ -0,0 +1,27 @@
1
+ {
2
+ "$mulmocast": {
3
+ "version": "1.1"
4
+ },
5
+ "lang": "en",
6
+ "title": "Test Deprecated OpenAI Image Models",
7
+ "imageParams": {
8
+ "provider": "openai",
9
+ "style": "Photorealistic-style"
10
+ },
11
+ "beats": [
12
+ {
13
+ "text": "dall-e-2 is deprecated; mulmocast rejects this with a migration hint before calling OpenAI",
14
+ "imagePrompt": "Blue sky, a flock of birds",
15
+ "imageParams": {
16
+ "model": "dall-e-2"
17
+ }
18
+ },
19
+ {
20
+ "text": "dall-e-3 is deprecated; mulmocast rejects this with a migration hint before calling OpenAI",
21
+ "imagePrompt": "Blue sky, a flock of birds",
22
+ "imageParams": {
23
+ "model": "dall-e-3"
24
+ }
25
+ }
26
+ ]
27
+ }
@@ -12,7 +12,7 @@
12
12
  },
13
13
  "imageParams": {
14
14
  "provider": "openai",
15
- "model": "dall-e-3",
15
+ "model": "gpt-image-1-mini",
16
16
  "style": "Photo realistic, cinematic style.",
17
17
  "images": {
18
18
  "optimus": {
@@ -0,0 +1,26 @@
1
+ {
2
+ "$mulmocast": {
3
+ "version": "1.1"
4
+ },
5
+ "lang": "en",
6
+ "movieParams": {
7
+ "provider": "google"
8
+ },
9
+ "canvasSize": {
10
+ "width": 720,
11
+ "height": 1280
12
+ },
13
+ "imageParams": {
14
+ "provider": "openai",
15
+ "model": "dall-e-3",
16
+ "style": "Photo realistic, cinematic style."
17
+ },
18
+ "beats": [
19
+ {
20
+ "text": "dall-e-3 on movie flow is deprecated; mulmocast rejects upfront with a migration hint",
21
+ "imagePrompt": "A rocket on the launch pad at dusk.",
22
+ "moviePrompt": "Rocket lifts off.",
23
+ "duration": 5
24
+ }
25
+ ]
26
+ }
@@ -0,0 +1,79 @@
1
+ {
2
+ "$mulmocast": { "version": "1.1" },
3
+ "canvasSize": { "width": 1024, "height": 1024 },
4
+ "imageParams": {
5
+ "provider": "replicate"
6
+ },
7
+ "speechParams": {
8
+ "speakers": {
9
+ "Presenter": {
10
+ "voiceId": "shimmer",
11
+ "provider": "openai"
12
+ }
13
+ }
14
+ },
15
+ "audioParams": { "bgmVolume": 0 },
16
+ "lang": "en",
17
+ "title": "Replicate Image Models Showcase",
18
+ "beats": [
19
+ {
20
+ "speaker": "Presenter",
21
+ "text": "PROMPT: A red panda holding a tiny espresso cup, soft window light, cinematic.",
22
+ "image": {
23
+ "type": "textSlide",
24
+ "slide": {
25
+ "title": "PROMPT: A red panda holding a tiny espresso cup, soft window light, cinematic."
26
+ }
27
+ }
28
+ },
29
+ {
30
+ "id": "flux-2-pro",
31
+ "speaker": "Presenter",
32
+ "text": "Flux 2 Pro from Black Forest Labs — high quality text-to-image.",
33
+ "imagePrompt": "A red panda holding a tiny espresso cup, soft window light, cinematic",
34
+ "imageParams": { "model": "black-forest-labs/flux-2-pro" }
35
+ },
36
+ {
37
+ "id": "flux-1.1-pro-ultra",
38
+ "speaker": "Presenter",
39
+ "text": "Flux 1.1 Pro Ultra — ultra-high resolution variant.",
40
+ "imagePrompt": "A red panda holding a tiny espresso cup, soft window light, cinematic",
41
+ "imageParams": { "model": "black-forest-labs/flux-1.1-pro-ultra" }
42
+ },
43
+ {
44
+ "id": "flux-schnell",
45
+ "speaker": "Presenter",
46
+ "text": "Flux Schnell — fastest open Flux model.",
47
+ "imagePrompt": "A red panda holding a tiny espresso cup, soft window light, cinematic",
48
+ "imageParams": { "model": "black-forest-labs/flux-schnell" }
49
+ },
50
+ {
51
+ "id": "ideogram-v3-turbo",
52
+ "speaker": "Presenter",
53
+ "text": "Ideogram v3 Turbo — strong at typography and posters.",
54
+ "imagePrompt": "A red panda holding a tiny espresso cup, soft window light, cinematic",
55
+ "imageParams": { "model": "ideogram-ai/ideogram-v3-turbo" }
56
+ },
57
+ {
58
+ "id": "recraft-v3",
59
+ "speaker": "Presenter",
60
+ "text": "Recraft v3 — strong at clean illustration and text in images.",
61
+ "imagePrompt": "A red panda holding a tiny espresso cup, soft window light, cinematic",
62
+ "imageParams": { "model": "recraft-ai/recraft-v3" }
63
+ },
64
+ {
65
+ "id": "sd-3.5-large",
66
+ "speaker": "Presenter",
67
+ "text": "Stable Diffusion 3.5 Large — open weights baseline.",
68
+ "imagePrompt": "A red panda holding a tiny espresso cup, soft window light, cinematic",
69
+ "imageParams": { "model": "stability-ai/stable-diffusion-3.5-large" }
70
+ },
71
+ {
72
+ "id": "luma-photon",
73
+ "speaker": "Presenter",
74
+ "text": "Luma Photon — strong photorealism and style transfer.",
75
+ "imagePrompt": "A red panda holding a tiny espresso cup, soft window light, cinematic",
76
+ "imageParams": { "model": "luma/photon" }
77
+ }
78
+ ]
79
+ }
@@ -0,0 +1,50 @@
1
+ {
2
+ "$mulmocast": { "version": "1.1" },
3
+ "movieParams": {
4
+ "provider": "replicate"
5
+ },
6
+ "audioParams": { "bgmVolume": 0 },
7
+ "captionParams": { "lang": "en" },
8
+ "lang": "en",
9
+ "title": "Replicate New Movie Models Showcase",
10
+ "beats": [
11
+ {
12
+ "text": "PROMPT: a woman is walking through a busy Tokyo street at night, she is wearing dark sunglasses",
13
+ "image": {
14
+ "type": "textSlide",
15
+ "slide": {
16
+ "title": "PROMPT: a woman is walking through a busy Tokyo street at night, she is wearing dark sunglasses"
17
+ }
18
+ }
19
+ },
20
+ {
21
+ "id": "alibaba-happyhorse",
22
+ "text": "alibaba/happyhorse-1.0",
23
+ "duration": 5,
24
+ "moviePrompt": "a woman is walking through a busy Tokyo street at night, she is wearing dark sunglasses",
25
+ "movieParams": { "model": "alibaba/happyhorse-1.0" }
26
+ },
27
+ {
28
+ "id": "hailuo-23",
29
+ "text": "minimax/hailuo-2.3",
30
+ "duration": 6,
31
+ "moviePrompt": "a woman is walking through a busy Tokyo street at night, she is wearing dark sunglasses",
32
+ "movieParams": { "model": "minimax/hailuo-2.3" }
33
+ },
34
+ {
35
+ "id": "pixverse-v5",
36
+ "text": "pixverse/pixverse-v5",
37
+ "duration": 5,
38
+ "moviePrompt": "a woman is walking through a busy Tokyo street at night, she is wearing dark sunglasses",
39
+ "movieParams": { "model": "pixverse/pixverse-v5" }
40
+ },
41
+ {
42
+ "id": "hailuo-23-fast",
43
+ "text": "minimax/hailuo-2.3-fast (i2v: starts from a generated still image)",
44
+ "duration": 6,
45
+ "imagePrompt": "a woman wearing dark sunglasses standing on a busy Tokyo street at night, neon reflections, cinematic still",
46
+ "moviePrompt": "the woman starts walking forward, neon reflections shimmer on the wet pavement",
47
+ "movieParams": { "model": "minimax/hailuo-2.3-fast" }
48
+ }
49
+ ]
50
+ }