@picsart/ai-sdk 6.11.1 → 6.13.0

package/index.d.ts

@@ -204,6 +204,16 @@ type ModelInputById = {
         styleExaggeration?: number;
         speed?: number;
         useSpeakerBoost?: boolean;
+        withTimestamps?: boolean;
+    };
+    "eleven-speech-to-text": {
+        audioUrl: string;
+        language?: string;
+        diarize?: boolean;
+        numSpeakers?: number;
+        timestampsGranularity?: "word" | "character";
+        tagAudioEvents?: boolean;
+        seed?: number;
     };
     "eleven-sts-v2": {
         audioUrl: string;
@@ -232,6 +242,11 @@ type ModelInputById = {
         styleExaggeration?: number;
         speed?: number;
         useSpeakerBoost?: boolean;
+        withTimestamps?: boolean;
+    };
+    "eleven-video-to-music": {
+        videoUrls: [string, ...string[]];
+        prompt?: string;
     };
     "eleven-voice-create": {
         prompt: string;
@@ -390,6 +405,18 @@ type ModelInputById = {
         imageUrls?: string[];
         thinking?: "off" | "low" | "medium" | "high";
     };
+    "gemini-3.8-flash-lite-tts": {
+        language?: string;
+        accent?: string;
+        prompt: string;
+        voiceId?: string;
+    };
+    "gemini-3.8-flash-tts": {
+        language?: string;
+        accent?: string;
+        prompt: string;
+        voiceId?: string;
+    };
     "gemini-omni-1.1-flash-preview": {
         prompt: string;
         aspectRatio?: "16:9" | "9:16";
@@ -1725,6 +1752,13 @@ type ModelInputById = {
         imageUrls?: string[];
         negativePrompt?: string;
     };
+    "seedream-5.0-flash": {
+        resolution?: "1K" | "2K";
+        prompt: string;
+        aspectRatio?: "1:1" | "4:3" | "3:4" | "16:9" | "9:16" | "3:2" | "2:3" | "21:9";
+        imageUrls?: string[];
+        negativePrompt?: string;
+    };
     "seedream-5.0-lite": {
         resolution?: "2K" | "3K";
         prompt: string;
@@ -1885,7 +1919,7 @@ type ModelInputById = {
 type TypedModelId = keyof ModelInputById;
 type ModelInput<M extends TypedModelId> = ModelInputById[M];
 /** IDs of text-generation (LLM) models — narrows generateText(). */
-type TextModelId = "claude-fable-5" | "claude-fable-5-1" | "claude-haiku-4-5" | "claude-opus-4-8" | "claude-opus-5" | "claude-sonnet-4-5" | "claude-sonnet-4-6" | "claude-sonnet-5" | "gemini-2.5-flash" | "gemini-3-pro" | "gemini-3.5-flash-lite" | "gemini-3.6-flash" | "gemini-3.7-flash" | "gemini-3.8-flash" | "gpt-4.1-mini" | "gpt-4.1-nano" | "gpt-4o" | "gpt-4o-mini" | "gpt-5" | "gpt-5-mini" | "gpt-5.1" | "gpt-5.2" | "gpt-5.5" | "gpt-5.6-luna" | "gpt-5.6-sol" | "gpt-5.6-terra" | "gpt-6-astra";
+type TextModelId = "claude-fable-5" | "claude-fable-5-1" | "claude-haiku-4-5" | "claude-opus-4-8" | "claude-opus-5" | "claude-sonnet-4-5" | "claude-sonnet-4-6" | "claude-sonnet-5" | "eleven-speech-to-text" | "gemini-2.5-flash" | "gemini-3-pro" | "gemini-3.5-flash-lite" | "gemini-3.6-flash" | "gemini-3.7-flash" | "gemini-3.8-flash" | "gpt-4.1-mini" | "gpt-4.1-nano" | "gpt-4o" | "gpt-4o-mini" | "gpt-5" | "gpt-5-mini" | "gpt-5.1" | "gpt-5.2" | "gpt-5.5" | "gpt-5.6-luna" | "gpt-5.6-sol" | "gpt-5.6-terra" | "gpt-6-astra";
 type TextModelInputById = Pick<ModelInputById, TextModelId>;
 
 /**
@@ -2423,7 +2457,7 @@ type Provider = AppProvider;
 /** App generation modes. */
 type GenerationMode = 'video' | 'image' | 'audio' | 'text';
 /** App input types. */
-type InputType = 't2v' | 'i2v' | 'v2v' | 'a2v' | 't2i' | 'i2i' | 't2a' | 'v2a' | 'tts' | 'sts' | 'sfx' | 'music' | 't2t' | 'i2t' | 'v2t';
+type InputType = 't2v' | 'i2v' | 'v2v' | 'a2v' | 't2i' | 'i2i' | 't2a' | 'v2a' | 'tts' | 'sts' | 'sfx' | 'music' | 't2t' | 'i2t' | 'v2t' | 'a2t';
 interface ModelFeature {
     label: string;
     variant: 'frame' | 'resolution' | 'audio' | 'duration' | 'input' | 'quality' | 'style' | 'characteristic';
@@ -3083,9 +3117,11 @@ declare const Models: {
     readonly ElevenDubbing: "eleven-dubbing";
     readonly ElevenMultilingualStsV2: "eleven-multilingual-sts-v2";
     readonly ElevenMultilingualV2: "eleven-multilingual-v2";
+    readonly ElevenSpeechToText: "eleven-speech-to-text";
     readonly ElevenStsV2: "eleven-sts-v2";
     readonly ElevenTextToDialogue: "eleven-text-to-dialogue";
     readonly ElevenV3: "eleven-v3";
+    readonly ElevenVideoToMusic: "eleven-video-to-music";
     readonly ElevenVoiceCreate: "eleven-voice-create";
     readonly ElevenVoiceDesignV2: "eleven-voice-design-v2";
     readonly ElevenVoiceDesignV3: "eleven-voice-design-v3";
@@ -3113,6 +3149,8 @@ declare const Models: {
     readonly Gemini36Flash: "gemini-3.6-flash";
     readonly Gemini37Flash: "gemini-3.7-flash";
     readonly Gemini38Flash: "gemini-3.8-flash";
+    readonly Gemini38FlashLiteTts: "gemini-3.8-flash-lite-tts";
+    readonly Gemini38FlashTts: "gemini-3.8-flash-tts";
     readonly GeminiOmni11FlashPreview: "gemini-omni-1.1-flash-preview";
     readonly GeminiOmniFlashPreview: "gemini-omni-flash-preview";
     readonly Gpt41Mini: "gpt-4.1-mini";
@@ -3290,6 +3328,7 @@ declare const Models: {
     readonly Seedream40: "seedream-4.0";
     readonly Seedream45: "seedream-4.5";
     readonly Seedream47: "seedream-4.7";
+    readonly Seedream50Flash: "seedream-5.0-flash";
     readonly Seedream50Lite: "seedream-5.0-lite";
     readonly Seedream50Pro: "seedream-5.0-pro";
     readonly TopazUpscaleImage: "topaz-upscale-image";

package/index.js

@@ -5070,6 +5070,7 @@ var buildSeedream45Payload = buildSeedreamV2("seedream_4_5");
 var buildSeedream47Payload = buildSeedreamV2("seedream_4_7");
 var buildSeedream50LitePayload = buildSeedreamV2("seedream_5_0_lite");
 var buildSeedream50ProPayload = buildSeedreamV2("seedream_5_0_pro");
+var buildSeedream50FlashPayload = buildSeedreamV2("seedream_5_0_flash");
 var seedreamV2Params = {
   ...params.prompt(),
   ...params.aspectRatio(["1:1", "4:3", "3:4", "16:9", "9:16", "3:2", "2:3", "21:9"], "16:9"),
@@ -5078,6 +5079,29 @@ var seedreamV2Params = {
   ...params.negativePrompt()
 };
 var { MODELS: MODELS14 } = defineModels("seedream", [
+  {
+    id: "seedream-5.0-flash",
+    name: "Seedream 5.0 Flash",
+    modelId: "seedream_5_0_flash",
+    addedAt: "2026-09-24",
+    workflow: "seedream",
+    buildPayload: buildSeedream50FlashPayload,
+    estimatedTime: { "1K": 10, "2K": 18 },
+    mode: "image",
+    inputType: "t2i",
+    // Fastest 5.0 tier. Backend gates Flash to 1K/2K (same as 5.0-pro). Single-image
+    // only (no group/sequential) with up to 10 reference images. Supports both T2I
+    // and I2I modes.
+    description: "Fastest 5.0 tier \u2014 quick 2K generation with up to 10 reference images.",
+    features: [feat("Multi-Image Input", "input"), feat("2K", "resolution")],
+    paramConfig: {
+      ...params.resolution(["1K", "2K"]),
+      ...params.prompt(),
+      ...params.aspectRatio(["1:1", "4:3", "3:4", "16:9", "9:16", "3:2", "2:3", "21:9"], "16:9"),
+      ...params.imageInput(10, "Source Images"),
+      ...params.negativePrompt()
+    }
+  },
   {
     id: "seedream-5.0-pro",
     name: "Seedream 5.0 Pro",
@@ -6383,9 +6407,9 @@ var buildGemini31FlashLiteImagePayload = (ctx) => ({
   imageSize: ctx.resolution ?? "1K",
   ...buildThinkingConfig(ctx)
 });
-var buildGeminiTTSPayload = (ctx) => ({
+var buildGeminiTTSPayload = (model) => (ctx) => ({
   text: ctx.prompt,
-  model: ctx.modelId ?? "gemini-2.5-flash-tts",
+  model,
   voiceName: ctx.voiceId ?? "Kore"
 });
 function inferMimeType2(url) {
@@ -6517,7 +6541,7 @@ var { MODELS: MODELS22 } = defineModels("google", [
     name: "Gemini 2.5 Flash TTS",
     addedAt: "2026-02-15",
     workflow: "gemini/v1/audios",
-    buildPayload: buildGeminiTTSPayload,
+    buildPayload: buildGeminiTTSPayload("gemini-2.5-flash-tts"),
     estimatedTime: 15,
     mode: "audio",
     inputType: "tts",
@@ -6537,7 +6561,7 @@ var { MODELS: MODELS22 } = defineModels("google", [
     name: "Gemini 2.5 Pro TTS",
     addedAt: "2026-03-18",
     workflow: "gemini/v1/audios",
-    buildPayload: buildGeminiTTSPayload,
+    buildPayload: buildGeminiTTSPayload("gemini-2.5-pro-tts"),
     estimatedTime: 20,
     mode: "audio",
     inputType: "tts",
@@ -6553,6 +6577,47 @@ var { MODELS: MODELS22 } = defineModels("google", [
       ...params.voiceId([], GEMINI_DEFAULT_VOICE_ID, { catalog: { workflow: "gemini/v1/catalog/voices" } })
     }
   },
+  {
+    id: "gemini-3.8-flash-tts",
+    name: "Gemini 3.8 Flash TTS",
+    addedAt: "2026-09-24",
+    workflow: "gemini/v1/audios",
+    buildPayload: buildGeminiTTSPayload("gemini-3.8-flash-tts"),
+    estimatedTime: 15,
+    mode: "audio",
+    inputType: "tts",
+    modelId: "gemini-3.8-flash-tts",
+    description: "Latest Gemini native text-to-speech with expressive multilingual voices.",
+    features: [feat("Multilingual", "characteristic"), feat("30 Voices", "characteristic")],
+    paramConfig: {
+      ...params.language(true),
+      // 6,000 boundary-verified against the live gemini/v1/audios worker
+      // (accepts >5,000; see scripts/api-tests/audio-charlimit-boundary-probe.mjs).
+      ...params.prompt({ maxLength: 6e3 }),
+      ...params.voiceId([], GEMINI_DEFAULT_VOICE_ID, { catalog: { workflow: "gemini/v1/catalog/voices" } })
+    }
+  },
+  {
+    id: "gemini-3.8-flash-lite-tts",
+    name: "Gemini 3.8 Flash Lite TTS",
+    addedAt: "2026-09-24",
+    workflow: "gemini/v1/audios",
+    buildPayload: buildGeminiTTSPayload("gemini-3.8-flash-lite-tts"),
+    estimatedTime: 12,
+    mode: "audio",
+    inputType: "tts",
+    modelId: "gemini-3.8-flash-lite-tts",
+    badge: ["fast"],
+    description: "Lightweight Gemini 3.8 TTS variant for faster, high-volume speech synthesis.",
+    features: [feat("Multilingual", "characteristic"), feat("30 Voices", "characteristic")],
+    paramConfig: {
+      ...params.language(true),
+      // 6,000 boundary-verified against the live gemini/v1/audios worker
+      // (accepts >5,000; see scripts/api-tests/audio-charlimit-boundary-probe.mjs).
+      ...params.prompt({ maxLength: 6e3 }),
+      ...params.voiceId([], GEMINI_DEFAULT_VOICE_ID, { catalog: { workflow: "gemini/v1/catalog/voices" } })
+    }
+  },
   // ── Video ─────────────────────────────────────────────────────────
   {
     id: "gemini-omni-flash-preview",
@@ -6937,7 +7002,11 @@ var ttsParamConfig = (promptMaxLength, withLanguage) => ({
   ...withLanguage ? params.language(false) : {},
   ...params.prompt({ maxLength: promptMaxLength }),
   ...params.voiceId([], DEFAULT_VOICE_ID, { catalog: VOICE_CATALOG }),
-  ...voiceSettingsParams(true)
+  ...voiceSettingsParams(true),
+  // Timings come back beside the audio, for captions and lip sync. Off by
+  // default: the vendor answers a different route that inlines the audio as
+  // base64, and callers who do not need timings should not pay the decode.
+  ...p.boolean("withTimestamps", false, "Character Timings")
 });
 var { MODELS: MODELS24 } = defineModels("elevenlabs", [
   // ── TTS ───────────────────────────────────────────────────────────
@@ -7063,6 +7132,56 @@ var { MODELS: MODELS24 } = defineModels("elevenlabs", [
       ...p.boolean("isInstrumental", false, "Instrumental Only")
     }
   },
+  // ── Transcription ─────────────────────────────────────────────────
+  {
+    // `modelId` is the pricing-catalog key: the vendor id `scribe_v2` is not
+    // vendor-scoped, so pricing carries the `eleven_` prefix, as music does.
+    id: "eleven-speech-to-text",
+    name: "Eleven Scribe v2",
+    modelId: "eleven_scribe_v2",
+    addedAt: "2026-09-24",
+    workflow: "elevenlabs/v1/speech-to-text",
+    estimatedTime: 10,
+    mode: "text",
+    inputType: "a2t",
+    description: "Transcribe speech from audio or video, with word timings and speaker labels.",
+    features: [feat("Transcription", "characteristic"), feat("Speaker Labels", "characteristic")],
+    paramConfig: {
+      ...params.audioInput("Audio or Video", true),
+      // ISO 639-1 or 639-3. Left free text: the vendor detects the language on
+      // its own, and the closed list would be 99 entries of upkeep.
+      language: {
+        label: "Language (ISO code, optional)",
+        descriptor: { kind: "text", placeholder: "e.g. en, rus \u2014 omit to auto-detect" }
+      },
+      ...p.boolean("diarize", false, "Label Speakers"),
+      ...p.range("numSpeakers", 1, 32, 1, { step: 1, label: "Speakers" }),
+      ...p.enum("timestampsGranularity", ["word", "character"], "word", { label: "Timing Detail" }),
+      ...p.boolean("tagAudioEvents", false, "Tag Audio Events"),
+      ...params.seed(2147483647)
+    }
+  },
+  // ── Video to Music ────────────────────────────────────────────────
+  {
+    // Same pricing model as music-generation — the same engine writes the
+    // track; what separates the two is the use case, `video-to-audio`.
+    id: "eleven-video-to-music",
+    name: "Eleven Video to Music",
+    modelId: "eleven_music_v2",
+    addedAt: "2026-09-24",
+    workflow: "elevenlabs/v1/video-to-music",
+    estimatedTime: 25,
+    mode: "audio",
+    inputType: "v2a",
+    description: "Score a video with a soundtrack written to follow what happens on screen.",
+    features: [feat("Video Scoring", "characteristic"), feat("Soundtrack", "characteristic")],
+    paramConfig: {
+      // The vendor takes up to ten clips, 200MB and 600 seconds in total, and
+      // scores them as one timeline.
+      ...params.videoInputs(10, "Videos", true),
+      ...params.prompt({ maxLength: 1e3, required: false, placeholder: "How the soundtrack should sound" })
+    }
+  },
   // ── Speech-to-Speech ──────────────────────────────────────────────
   {
     id: "eleven-sts-v2",
@@ -7227,6 +7346,7 @@ var buildElevenLabsTTSPayload = (modelId) => (input) => ({
   voice_id: input.voiceId ?? DEFAULT_VOICE_ID,
   model_id: modelId,
   ...input.language ? { language_code: input.language } : {},
+  ...input.withTimestamps ? { with_timestamps: true } : {},
   ...voiceSettings(input)
 });
 var buildElevenLabsSTSPayload = (modelId) => (input) => ({
@@ -7246,13 +7366,30 @@ var buildElevenLabsDialoguePayload = (input) => ({
   ...input.language ? { language_code: input.language } : {},
   ...input.seed != null ? { seed: input.seed } : {}
 });
+var buildElevenLabsTranscribePayload = (input) => ({
+  audio_url: input.audioUrl,
+  ...input.language ? { language_code: input.language } : {},
+  ...input.diarize ? { diarize: true } : {},
+  // Only meaningful alongside diarization, and only when the caller moved it
+  // off the default of one.
+  ...input.diarize && input.numSpeakers && input.numSpeakers > 1 ? { num_speakers: input.numSpeakers } : {},
+  ...input.timestampsGranularity ? { timestamps_granularity: input.timestampsGranularity } : {},
+  ...input.tagAudioEvents ? { tag_audio_events: true } : {},
+  ...input.seed != null ? { seed: input.seed } : {}
+});
+var buildElevenLabsVideoToMusicPayload = (input) => ({
+  video_urls: input.videoUrls,
+  ...input.prompt ? { description: input.prompt } : {}
+});
 registerPayloads(MODELS24, {
   "elevenlabs-music-v2": buildElevenLabsMusicPayload,
   "eleven-v3": buildElevenLabsTTSPayload("eleven_v3"),
   "eleven-multilingual-v2": buildElevenLabsTTSPayload("eleven_multilingual_v2"),
   "eleven-sts-v2": buildElevenLabsSTSPayload("eleven_english_sts_v2"),
   "eleven-multilingual-sts-v2": buildElevenLabsSTSPayload("eleven_multilingual_sts_v2"),
-  "eleven-text-to-dialogue": buildElevenLabsDialoguePayload
+  "eleven-text-to-dialogue": buildElevenLabsDialoguePayload,
+  "eleven-speech-to-text": buildElevenLabsTranscribePayload,
+  "eleven-video-to-music": buildElevenLabsVideoToMusicPayload
 });
 
 // src/vendors/catalog/heygen.ts
@@ -12587,9 +12724,11 @@ var ElevenAudioIsolation = "eleven-audio-isolation";
 var ElevenDubbing = "eleven-dubbing";
 var ElevenMultilingualStsV2 = "eleven-multilingual-sts-v2";
 var ElevenMultilingualV2 = "eleven-multilingual-v2";
+var ElevenSpeechToText = "eleven-speech-to-text";
 var ElevenStsV2 = "eleven-sts-v2";
 var ElevenTextToDialogue = "eleven-text-to-dialogue";
 var ElevenV3 = "eleven-v3";
+var ElevenVideoToMusic = "eleven-video-to-music";
 var ElevenVoiceCreate = "eleven-voice-create";
 var ElevenVoiceDesignV2 = "eleven-voice-design-v2";
 var ElevenVoiceDesignV3 = "eleven-voice-design-v3";
@@ -12617,6 +12756,8 @@ var Gemini35FlashLite = "gemini-3.5-flash-lite";
 var Gemini36Flash = "gemini-3.6-flash";
 var Gemini37Flash = "gemini-3.7-flash";
 var Gemini38Flash = "gemini-3.8-flash";
+var Gemini38FlashLiteTts = "gemini-3.8-flash-lite-tts";
+var Gemini38FlashTts = "gemini-3.8-flash-tts";
 var GeminiOmni11FlashPreview = "gemini-omni-1.1-flash-preview";
 var GeminiOmniFlashPreview = "gemini-omni-flash-preview";
 var Gpt41Mini = "gpt-4.1-mini";
@@ -12794,6 +12935,7 @@ var SeedanceI2v = "seedance-i2v";
 var Seedream40 = "seedream-4.0";
 var Seedream45 = "seedream-4.5";
 var Seedream47 = "seedream-4.7";
+var Seedream50Flash = "seedream-5.0-flash";
 var Seedream50Lite = "seedream-5.0-lite";
 var Seedream50Pro = "seedream-5.0-pro";
 var TopazUpscaleImage = "topaz-upscale-image";
@@ -12832,9 +12974,11 @@ var Models = {
   ElevenDubbing,
   ElevenMultilingualStsV2,
   ElevenMultilingualV2,
+  ElevenSpeechToText,
   ElevenStsV2,
   ElevenTextToDialogue,
   ElevenV3,
+  ElevenVideoToMusic,
   ElevenVoiceCreate,
   ElevenVoiceDesignV2,
   ElevenVoiceDesignV3,
@@ -12862,6 +13006,8 @@ var Models = {
   Gemini36Flash,
   Gemini37Flash,
   Gemini38Flash,
+  Gemini38FlashLiteTts,
+  Gemini38FlashTts,
   GeminiOmni11FlashPreview,
   GeminiOmniFlashPreview,
   Gpt41Mini,
@@ -13039,6 +13185,7 @@ var Models = {
   Seedream40,
   Seedream45,
   Seedream47,
+  Seedream50Flash,
   Seedream50Lite,
   Seedream50Pro,
   TopazUpscaleImage,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@picsart/ai-sdk",
-  "version": "6.11.1",
+  "version": "6.13.0",
   "type": "module",
   "description": "Type-safe SDK for 100+ AI models — image, video, audio, and text generation with Picsart",
   "license": "MIT",