samsar-js 0.48.32 → 0.48.33

package/README.md

@@ -37,6 +37,10 @@ const video = await samsar.createVideoFromText(
     image_model: 'GPTIMAGE2',
     video_model: 'RUNWAYML',
     duration: 30,
+    backingtrack_model: 'LYRIA3',
+    tts_model: 'OPENAI',
+    speakerOptions: { openAISpeakers: ['nova'] },
+    inference_model: 'gpt-5.5',
     font_key: 'Poppins',
     enable_subtitles: true,
   },
@@ -67,6 +71,10 @@ const videoFromImages = await samsar.createVideoFromImageList(
     video_model: 'RUNWAYML',
     aspect_ratio: '16:9',
     language: 'en',
+    backingtrack_model: 'ELEVENLABS_MUSIC',
+    tts_model: 'ELEVENLABS',
+    speakerOptions: { elevenLabsSpeakers: ['EXAVITQu4vr4xnSDxMaL'] },
+    inference_model: 'gemini-3.1-pro',
     font_key: 'Poppins',
     enable_subtitles: true,
   },
@@ -574,9 +582,11 @@ console.log(externalLibrary.data.requests.map((request) => request.request_id));
 
 Video model support notes:
 - `createVideoFromText` image model keys include: `GPTIMAGE2`, `IMAGEN4`, `SEEDREAM`, `NANOBANANA2`, `NANOBANANAPRO`, `CUSTOM_TEXT_TO_IMAGE`.
-- `createVideoFromText` supports these video models: `RUNWAYML`, `VEO3.1I2V`, `VEO3.1I2VFAST`, `SEEDANCEI2V` (Seedance 1.5), `KLINGIMGTOVID3PRO`, and `HAPPYHORSEI2V`.
+- `createVideoFromText` supports these video models: `RUNWAYML`, `VEO3.1I2V`, `VEO3.1I2VFAST`, `COSMOS3SUPERI2V`, `SEEDANCEI2V` (Seedance 1.5), `KLINGIMGTOVID3PRO`, and `HAPPYHORSEI2V`.
 - `createVideoFromText` accepts either a provided outro (`outro_image_url`) or server-generated QR outro (`generate_outro_image: true` with `cta_url`). It can also render bottom CTA footer QR cards with `add_footer_animation` and `footer_metadata`; one footer item applies to every generated scene, while multiple items map by scene index.
-- `createVideoFromImageList` supports `RUNWAYML`, `VEO3.1I2V`, `VEO3.1I2VFAST`, `SEEDANCEI2V`, `KLINGIMGTOVID3PRO`, and `HAPPYHORSEI2V` via `video_model`; if omitted, it defaults to `RUNWAYML`. It also accepts the same `image_model` keys as text-to-video. Use `aspect_ratio: '16:9'` or `'9:16'`; omitted or invalid values fall back to `16:9`.
+- `createVideoFromImageList` supports `RUNWAYML`, `VEO3.1I2V`, `VEO3.1I2VFAST`, `COSMOS3SUPERI2V`, `SEEDANCEI2V`, `KLINGIMGTOVID3PRO`, and `HAPPYHORSEI2V` via `video_model`; if omitted, it defaults to `RUNWAYML`. It also accepts the same `image_model` keys as text-to-video. Use `aspect_ratio: '16:9'` or `'9:16'`; omitted or invalid values fall back to `16:9`.
+- Text and image-list video creation both accept optional `backingtrack_model` / `backing_track_model` / `backingTrackModel` / `music_provider` / `musicProvider`, `tts_model` / `ttsModel` / `tts_provider` / `ttsProvider`, `speakerOptions` / `speaker_options`, and `inference_model` / `inferenceModel`. The adapter normalizes these to `backingtrack_model`, `tts_model`, `speakerOptions`, and `inference_model` in the request payload. Omit `inference_model` to use the account default; supported request values are `gpt-5.5` and `gemini-3.1-pro`. When `tts_model` is set, Samsar limits assignment to the matching speaker list (`openAISpeakers`, `elevenLabsSpeakers`, or `googleSpeakers`; Google TTS requests should include `googleSpeakerDetails`).
+- `video_model_sub_type` is no longer used by the API and is stripped from text and image-list payloads before sending.
 - `createVideoFromImageList` accepts either a provided outro (`outro_image_url`) or server-generated QR outro (`generate_outro_image: true` with `cta_url`). Do not combine the two modes in a single request.
 - `createVideoFromImageList` can render per-scene footer QR cards by setting `add_footer_animation: true` and providing one `footer_metadata` item per image scene.
 - `createVideoFromImageList` can also generate QR outro CTA text and each scene footer CTA from a single link by setting `express_cta_generation: true` with `cta_url`. CamelCase `expressCtaGeneration` and compatibility aliases `auto_generate_cta_text` / `generate_cta_texts` are normalized to the same API field.

package/dist/index.d.ts

@@ -36,6 +36,36 @@ export interface FontOptions {
     language?: string;
     family?: string;
 }
+export type BackingTrackModel = 'LYRIA3' | 'LYRIA2' | 'ELEVENLABS_MUSIC' | 'Lyria 3' | 'Lyria 2' | 'ElevenLabs music' | string;
+export type TTSModel = 'ELEVENLABS' | 'OPENAI' | 'GOOGLE' | 'ElevenLabs' | 'OpenAI' | 'Google TTS' | string;
+export type InferenceModel = 'gpt-5.5' | 'GPT 5.5' | 'gemini-3.1-pro' | 'Gemini 3.1 Pro' | string;
+export interface GoogleTTSSpeakerDetail {
+    provider?: 'GOOGLE' | string;
+    value?: string;
+    voiceId?: string;
+    name?: string;
+    label?: string;
+    languageCode?: string;
+    languageCodes?: string[];
+    Gender?: 'M' | 'F' | '' | string | null;
+    gender?: string;
+    genderCode?: string;
+    ssmlGender?: string;
+    naturalSampleRateHertz?: number | null;
+    voiceType?: string;
+    previewRequiresAuth?: boolean;
+    [key: string]: unknown;
+}
+export interface TTSSpeakerOptions {
+    allowOpenAI?: boolean;
+    allowElevenLabs?: boolean;
+    allowGoogle?: boolean;
+    openAISpeakers?: string[];
+    elevenLabsSpeakers?: string[];
+    googleSpeakers?: string[];
+    googleSpeakerDetails?: GoogleTTSSpeakerDetail[];
+    [key: string]: unknown;
+}
 export type V2StepGenerationMode = 'one_step' | 'two_step';
 export interface V2StepGenerationOptions {
     auto_render_full_video?: boolean;
@@ -50,7 +80,19 @@ export interface CreateVideoFromTextInput extends V2StepGenerationOptions {
     duration: number;
     tone?: string;
     aspect_ratio?: string;
-    video_model_sub_type?: string;
+    backingtrack_model?: BackingTrackModel;
+    backing_track_model?: BackingTrackModel;
+    backingTrackModel?: BackingTrackModel;
+    music_provider?: BackingTrackModel;
+    musicProvider?: BackingTrackModel;
+    tts_model?: TTSModel;
+    ttsModel?: TTSModel;
+    tts_provider?: TTSModel;
+    ttsProvider?: TTSModel;
+    inference_model?: InferenceModel;
+    inferenceModel?: InferenceModel;
+    speakerOptions?: TTSSpeakerOptions;
+    speaker_options?: TTSSpeakerOptions;
     font_key?: string;
     fontKey?: string;
     subtitle_font?: string;
@@ -118,7 +160,7 @@ export interface FooterMetadataItem {
     footerLogoImagePath?: string;
 }
 export type ImageListToVideoAspectRatio = '16:9' | '9:16';
-export type ImageListToVideoModel = 'RUNWAYML' | 'VEO3.1I2V' | 'VEO3.1I2VFAST' | 'SEEDANCEI2V' | 'KLINGIMGTOVID3PRO' | 'HAPPYHORSEI2V';
+export type ImageListToVideoModel = 'RUNWAYML' | 'VEO3.1I2V' | 'VEO3.1I2VFAST' | 'COSMOS3SUPERI2V' | 'SEEDANCEI2V' | 'KLINGIMGTOVID3PRO' | 'HAPPYHORSEI2V';
 export interface ImageListToVideoItem {
     image_url?: string;
     imageUrl?: string;
@@ -155,6 +197,19 @@ export interface CreateVideoFromImageListInput extends V2StepGenerationOptions {
     aspectRatio?: ImageListToVideoAspectRatio;
     language?: string;
     languageString?: string | null;
+    backingtrack_model?: BackingTrackModel;
+    backing_track_model?: BackingTrackModel;
+    backingTrackModel?: BackingTrackModel;
+    music_provider?: BackingTrackModel;
+    musicProvider?: BackingTrackModel;
+    tts_model?: TTSModel;
+    ttsModel?: TTSModel;
+    tts_provider?: TTSModel;
+    ttsProvider?: TTSModel;
+    inference_model?: InferenceModel;
+    inferenceModel?: InferenceModel;
+    speakerOptions?: TTSSpeakerOptions;
+    speaker_options?: TTSSpeakerOptions;
     font_key?: string;
     fontKey?: string;
     subtitle_font?: string;

package/dist/index.js

@@ -137,6 +137,10 @@ function normalizeCreateVideoFromTextInput(input) {
         ['cta_logo', ['cta_logo', 'ctaLogo']],
         ['add_footer_animation', ['add_footer_animation', 'addFooterAnimation']],
         ['footer_metadata', ['footer_metadata', 'footerMetadata']],
+        ['backingtrack_model', ['backingtrack_model', 'backing_track_model', 'backingTrackModel', 'music_provider', 'musicProvider']],
+        ['tts_model', ['tts_model', 'ttsModel', 'tts_provider', 'ttsProvider']],
+        ['inference_model', ['inference_model', 'inferenceModel']],
+        ['speakerOptions', ['speakerOptions', 'speaker_options']],
         ['enable_subtitles', ['enable_subtitles', 'enableSubtitles']],
         ['font_key', ['font_key', 'fontKey']],
     ];
@@ -146,6 +150,8 @@ function normalizeCreateVideoFromTextInput(input) {
             normalized[canonicalName] = value;
         }
     }
+    delete normalized.video_model_sub_type;
+    delete normalized.videoModelSubType;
     assertOptionalBoolean(normalized.enable_subtitles, 'enable_subtitles', 'createVideoFromText');
     assertOptionalBoolean(normalized.add_outro_animation, 'add_outro_animation', 'createVideoFromText');
     assertOptionalBoolean(normalized.add_outro_focus_area, 'add_outro_focus_area', 'createVideoFromText');
@@ -193,6 +199,10 @@ function normalizeCreateVideoFromImageListInput(input) {
         ['cta_logo', ['cta_logo', 'ctaLogo']],
         ['add_footer_animation', ['add_footer_animation', 'addFooterAnimation']],
         ['footer_metadata', ['footer_metadata', 'footerMetadata']],
+        ['backingtrack_model', ['backingtrack_model', 'backing_track_model', 'backingTrackModel', 'music_provider', 'musicProvider']],
+        ['tts_model', ['tts_model', 'ttsModel', 'tts_provider', 'ttsProvider']],
+        ['inference_model', ['inference_model', 'inferenceModel']],
+        ['speakerOptions', ['speakerOptions', 'speaker_options']],
         ['limit_single_narrator', ['limit_single_narrator', 'limitSingleNarrator']],
         ['add_narrator_avatar', ['add_narrator_avatar', 'addNarratorAvatar']],
         ['enable_subtitles', ['enable_subtitles', 'enableSubtitles']],
@@ -204,6 +214,8 @@ function normalizeCreateVideoFromImageListInput(input) {
             normalized[canonicalName] = value;
         }
     }
+    delete normalized.video_model_sub_type;
+    delete normalized.videoModelSubType;
     assertOptionalBoolean(normalized.enable_subtitles, 'enable_subtitles');
     assertOptionalBoolean(normalized.add_outro_animation, 'add_outro_animation');
     assertOptionalBoolean(normalized.add_outro_focus_area, 'add_outro_focus_area');

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "samsar-js",
-  "version": "0.48.32",
+  "version": "0.48.33",
   "description": "TypeScript client for the Samsar Processor API routes.",
   "type": "module",
   "main": "dist/index.js",