@henjicc/ai-sdk 0.7.0 → 0.9.0

package/dist/catalog/apimart/gemini-omni-flash.model.js

@@ -86,12 +86,9 @@ export const apimartGeminiOmniFlashModel = defineModel({
                 { value: 'reference' },
                 { value: 'frame' }
             ]
-        },
-        {
-            id: 'apimartGeminiOmniFlashExtendTaskId', type: 'text', order: 5,
-            default: '',
-            visible: { condition: (params) => params.apimartGeminiOmniFlashChannel !== 'ext' }
         }
+        // extend_from_task_id（多轮延续）要填供应商侧的上一次任务 ID，界面不展示任务 ID、用户也拿不到：
+        // 按“不显示且不请求”处理（界面计划 5.8，P-17）；以后做成生成记录上的“继续编辑”动作再接入
     ],
     endpoints: '/v1/videos/generations',
     request: {
@@ -136,13 +133,7 @@ export const apimartGeminiOmniFlashModel = defineModel({
                     body.video_urls = videos.slice(0, 1);
                 return body;
             }
-            const extendTaskId = typeof params.apimartGeminiOmniFlashExtendTaskId === 'string'
-                ? params.apimartGeminiOmniFlashExtendTaskId.trim()
-                : '';
-            if (videos.length > 0 && extendTaskId) {
-                throw new Error('Gemini Omni Flash 官方渠道不能同时传入参考视频和延续任务 ID');
-            }
-            if (!prompt && images.length + videos.length === 0 && !extendTaskId) {
+            if (!prompt && images.length + videos.length === 0) {
                 throw new Error('Gemini Omni Flash 官方渠道至少需要提示词或一份参考素材');
             }
             const body = {
@@ -156,8 +147,6 @@ export const apimartGeminiOmniFlashModel = defineModel({
                 body.image_urls = images.slice(0, 4);
             if (videos.length > 0)
                 body.video_urls = videos.slice(0, 1);
-            if (extendTaskId)
-                body.extend_from_task_id = extendTaskId;
             return body;
         }
     },

package/dist/catalog/bailian/cosyvoice-v3.5.model.js

@@ -8,7 +8,7 @@ export const bailianCosyVoice35Model = defineModel({
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
         { id: 'bailianCosyVoiceSpec', type: 'dropdown', order: 1, default: 'flash', options: [{ value: 'flash' }, { value: 'plus' }] },
-        { id: 'bailianCosyVoiceId', type: 'text', order: 2, required: true, default: '' },
+        { id: 'bailianCosyVoiceId', type: 'composite', valueType: 'string', order: 2, required: true, default: '' },
         { id: 'bailianCosyVoiceInstruction', type: 'text', order: 3, default: '' },
         { id: 'bailianCosyVoiceSpeed', type: 'number', order: 4, default: 1, min: 0.5, max: 2, step: 0.1 },
     ],

package/dist/catalog/bailian/minimax-speech-2.8.model.js

@@ -8,7 +8,7 @@ export const bailianMinimaxSpeech28Model = defineModel({
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
         { id: 'bailianMinimaxSpeechSpec', type: 'dropdown', order: 1, default: 'hd', options: [{ value: 'hd' }, { value: 'turbo' }] },
-        { id: 'bailianMinimaxSpeechVoice', type: 'text', order: 2, default: 'male-qn-qingse' },
+        { id: 'bailianMinimaxSpeechVoice', type: 'composite', valueType: 'string', order: 2, default: 'male-qn-qingse' },
         { id: 'bailianMinimaxSpeechSpeed', type: 'number', order: 3, default: 1, min: 0.5, max: 2, step: 0.1 },
         { id: 'bailianMinimaxSpeechEmotion', type: 'dropdown', order: 4, default: '', options: ['', 'happy', 'sad', 'angry', 'fearful', 'disgusted', 'surprised', 'calm'].map(value => ({ value })) },
     ],

package/dist/catalog/bailian/qwen-audio-3.1-tts-flash.model.js

@@ -8,7 +8,7 @@ export const bailianQwenAudio31TtsFlashModel = defineModel({
     },
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
-        { id: 'bailianQwenTtsVoice', type: 'dropdown', order: 1, default: 'longanhuan_v3.1', options: ['longanhuan_v3.1', 'longanlingxin_v3.1', 'longanfengyue_v3.1', 'xunanchuan_v3.1'].map(value => ({ value })) },
+        { id: 'bailianQwenTtsVoice', type: 'composite', valueType: 'string', order: 1, default: 'longanhuan_v3.1' },
         { id: 'bailianQwenTtsInstruction', type: 'text', order: 2, default: '' },
         { id: 'bailianQwenTtsSpeed', type: 'number', order: 3, default: 1, min: 0.5, max: 2, step: 0.1 },
     ],

package/dist/catalog/fal/eleven-v3-tts.model.js

@@ -8,7 +8,7 @@ export const falElevenV3TtsModel = defineModel({
     },
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
-        { id: 'falElevenV3Voice', type: 'text', order: 1, default: 'Rachel' },
+        { id: 'falElevenV3Voice', type: 'composite', valueType: 'string', order: 1, default: 'Rachel' },
         { id: 'falElevenV3Stability', type: 'number', order: 2, default: 0.5, min: 0, max: 1, step: 0.05 },
         { id: 'falElevenV3Language', type: 'text', order: 3, default: '' },
     ],

package/dist/catalog/fal/minimax-speech-2.8.model.js

@@ -10,7 +10,7 @@ export const falMinimaxSpeech28Model = defineModel({
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
         { id: 'falMinimaxSpeechSpec', type: 'dropdown', order: 1, default: 'hd', options: [{ value: 'hd' }, { value: 'turbo' }] },
-        { id: 'falMinimaxSpeechVoice', type: 'text', order: 2, default: 'Wise_Woman' },
+        { id: 'falMinimaxSpeechVoice', type: 'composite', valueType: 'string', order: 2, default: 'Wise_Woman' },
         { id: 'falMinimaxSpeechSpeed', type: 'number', order: 3, default: 1, min: 0.5, max: 2, step: 0.1 },
         { id: 'falMinimaxSpeechEmotion', type: 'dropdown', order: 4, default: '', options: ['', 'happy', 'sad', 'angry', 'fearful', 'disgusted', 'surprised', 'neutral'].map(value => ({ value })) },
     ],

package/dist/catalog/kie/eleven-turbo-2.5-tts.model.js

@@ -8,7 +8,7 @@ export const kieElevenTurbo25TtsModel = defineModel({
     },
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
-        { id: 'kieElevenTurboVoice', type: 'text', order: 1, default: 'Rachel' },
+        { id: 'kieElevenTurboVoice', type: 'composite', valueType: 'string', order: 1, default: 'Rachel' },
         { id: 'kieElevenTurboStability', type: 'number', order: 2, default: 0.5, min: 0, max: 1, step: 0.05 },
         { id: 'kieElevenTurboSpeed', type: 'number', order: 3, default: 1, min: 0.7, max: 1.2, step: 0.05 },
     ],

package/dist/catalog/kie/eleven-v3-dialogue.model.js

@@ -20,7 +20,7 @@ export const kieElevenV3DialogueModel = defineModel({
     },
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
-        { id: 'kieElevenDialogueVoice', type: 'text', order: 1, default: 'EkK5I93UQWFDigLMpZcX' },
+        { id: 'kieElevenDialogueVoice', type: 'composite', valueType: 'string', order: 1, default: 'EkK5I93UQWFDigLMpZcX' },
     ],
     endpoints: '/api/v1/jobs/createTask',
     request: {

package/dist/catalog/kie/gemini-3.1-flash-tts.model.js

@@ -8,7 +8,7 @@ export const kieGemini31FlashTtsModel = defineModel({
     },
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
-        { id: 'kieGeminiTtsVoice', type: 'text', order: 1, default: 'Fenrir' },
+        { id: 'kieGeminiTtsVoice', type: 'composite', valueType: 'string', order: 1, default: 'Fenrir' },
         { id: 'kieGeminiTtsScene', type: 'text', order: 2, default: '' },
         { id: 'kieGeminiTtsTemperature', type: 'number', order: 3, default: 1, min: 0, max: 2, step: 0.1 },
     ],

package/dist/catalog/kie/gemini-omni-video.model.js

@@ -71,19 +71,9 @@ export const kieGeminiOmniVideoModel = defineModel({
                 { value: '1080p' },
                 { value: '4k' }
             ]
-        },
-        {
-            id: 'kieGeminiOmniVideoAudioIds',
-            type: 'textarea',
-            order: 4,
-            default: ''
-        },
-        {
-            id: 'kieGeminiOmniVideoCharacterIds',
-            type: 'textarea',
-            order: 5,
-            default: ''
         }
+        // audio_ids / character_ids 要先经 omni/audio、omni/character 两个端点创建资产才拿得到 ID，
+        // 本应用还没有资产创建与管理入口，用户填不出来：按“不显示且不请求”处理（界面计划 5.8，P-17）
     ],
     endpoints: KIE_CREATE_TASK_ENDPOINT,
     request: {
@@ -103,16 +93,9 @@ export const kieGeminiOmniVideoModel = defineModel({
             const images = pickSources(params.uploadedFilePaths, params.images);
             const videos = pickSources(params.uploadedVideoFilePaths, params.videos);
             const hasVideo = videos.length > 0;
-            const splitIds = (value) => typeof value === 'string'
-                ? value.split(/[\n,]/u).map((item) => item.trim()).filter(Boolean)
-                : [];
-            const audioIds = splitIds(params.kieGeminiOmniVideoAudioIds).slice(0, 3);
-            const characterIds = splitIds(params.kieGeminiOmniVideoCharacterIds);
-            if (hasVideo && characterIds.length > 3)
-                throw new Error('Gemini Omni 带视频时最多支持 3 个角色资产');
-            const availableImageSlots = (hasVideo ? MAX_IMAGES_WITH_VIDEO : MAX_IMAGES_NO_VIDEO) - characterIds.length;
-            if (availableImageSlots < 0 || images.length > availableImageSlots) {
-                throw new Error('Gemini Omni 的图片与角色资产超过共享的 7 个槽位');
+            const availableImageSlots = hasVideo ? MAX_IMAGES_WITH_VIDEO : MAX_IMAGES_NO_VIDEO;
+            if (images.length > availableImageSlots) {
+                throw new Error(`Gemini Omni 最多支持 ${availableImageSlots} 张参考图（共享 7 个槽位，参考视频占 2 个）`);
             }
             const resolution = params.kieGeminiOmniVideoResolution === '1080p' || params.kieGeminiOmniVideoResolution === '4k'
                 ? params.kieGeminiOmniVideoResolution
@@ -149,10 +132,6 @@ export const kieGeminiOmniVideoModel = defineModel({
             if (images.length > 0) {
                 input.image_urls = images;
             }
-            if (audioIds.length > 0)
-                input.audio_ids = audioIds;
-            if (characterIds.length > 0)
-                input.character_ids = characterIds;
             if (hasVideo) {
                 const rawClipDuration = typeof params.__firstVideoDurationSeconds === 'number' && params.__firstVideoDurationSeconds > 0
                     ? params.__firstVideoDurationSeconds

package/dist/catalog/ppio/minimax-speech.model.js

@@ -11,7 +11,7 @@ export const minimaxSpeechModel = defineModel({
         canonicalModelId: 'minimax-speech-2.8',
         provider: 'ppio',
         type: 'audio',
-        tags: ['audio', 'text-to-speech', 'voice-synthesis', 'voice-cloning', 'provider-ppio'],
+        tags: ['audio', 'text-to-speech', 'voice-synthesis', 'voice-cloning', 'voice-clone', 'provider-ppio'],
         polling: {
             interval: 3000,
             maxAttempts: 120,

package/dist/catalog/siliconflow/cosyvoice2-tts.model.js

@@ -7,7 +7,7 @@ export const siliconflowCosyVoice2TtsModel = defineModel({
     },
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
-        { id: 'siliconflowCosyVoice', type: 'text', order: 1, default: 'FunAudioLLM/CosyVoice2-0.5B:alex' },
+        { id: 'siliconflowCosyVoice', type: 'composite', valueType: 'string', order: 1, default: 'FunAudioLLM/CosyVoice2-0.5B:alex' },
         { id: 'siliconflowCosySpeed', type: 'number', order: 2, default: 1, min: 0.25, max: 4, step: 0.05 },
         { id: 'siliconflowCosyFormat', type: 'dropdown', order: 3, default: 'mp3', options: ['mp3', 'opus', 'wav', 'pcm'].map(value => ({ value })) },
     ],

package/dist/catalog/siliconflow/moss-ttsd-0.5.model.js

@@ -7,7 +7,7 @@ export const siliconflowMossTtsd05Model = defineModel({
     },
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
-        { id: 'siliconflowMossVoice', type: 'text', order: 1, default: 'fnlp/MOSS-TTSD-v0.5:alex' },
+        { id: 'siliconflowMossVoice', type: 'composite', valueType: 'string', order: 1, default: 'fnlp/MOSS-TTSD-v0.5:alex' },
         { id: 'siliconflowMossFormat', type: 'dropdown', order: 2, default: 'mp3', options: ['mp3', 'opus', 'wav', 'pcm'].map(value => ({ value })) },
     ],
     endpoints: '/v1/audio/speech',

package/dist/catalog/volcengine-speech/seed-icl-2.0.model.d.ts

@@ -1,2 +1,3 @@
+export declare const VOLC_CLONE_LANGUAGES: number[];
 export declare const volcengineSeedIcl20Model: import("../../index.js").ModelRuntimeDefinition;
 export default volcengineSeedIcl20Model;

package/dist/catalog/volcengine-speech/seed-icl-2.0.model.js

@@ -1,26 +1,68 @@
 import { defineModel } from '../defineModel.js';
-import { speechString, speechText } from '../audio/tts-common.js';
+import { speechCharacterPrice, speechString, speechText } from '../audio/tts-common.js';
+import { DEFAULT_SEED_TTS_20_VOICE, isSeedTts20SystemVoice } from './seed-tts-2.0-voices.js';
+const cloning = (params) => params.volcIclMode === 'clone';
+const synth = (params) => !cloning(params);
+export const VOLC_CLONE_LANGUAGES = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 19, 20, 21];
 export const volcengineSeedIcl20Model = defineModel({
     meta: {
         id: 'volcengine-seed-icl-2.0', canonicalModelId: 'seed-icl-2.0', provider: 'volcengine-speech', type: 'audio',
-        tags: ['text-to-audio', 'provider-volcengine'],
+        tags: ['text-to-audio', 'voice-clone', 'provider-volcengine'],
+        polling: { interval: 3000, maxAttempts: 120 },
     },
     inputLimits: { images: { max: 0 }, videos: { max: 0 } },
     params: [
-        { id: 'volcSeedIclSpeaker', type: 'text', required: true, order: 1, default: '' },
+        { id: 'volcIclMode', type: 'dropdown', order: 1, default: 'speech', options: [{ value: 'speech' }, { value: 'clone' }] },
+        { id: 'volcSeedIclSpeaker', type: 'composite', valueType: 'string', required: true, order: 2, default: DEFAULT_SEED_TTS_20_VOICE, visible: { condition: synth } },
+        { id: 'volcIclActivationConsent', type: 'switch', order: 3, default: false, visible: { condition: p => synth(p) && Boolean(p.volcSeedIclSpeaker) && !isSeedTts20SystemVoice(String(p.volcSeedIclSpeaker)) && !/^S_/i.test(String(p.volcSeedIclSpeaker)) } },
+        { id: 'volcCloneName', type: 'text', order: 4, default: '', maxLength: 80, required: true, visible: { condition: cloning } },
+        { id: 'volcCloneAudio', type: 'file-upload', order: 5, default: [], maxCount: 1, maxSize: 10_000_000, accept: ['audio/wav', 'audio/x-wav', 'audio/mpeg', 'audio/ogg', 'audio/mp4', 'audio/x-m4a', 'audio/aac'], required: true, visible: { condition: cloning } },
+        { id: 'volcCloneLanguage', type: 'dropdown', order: 6, default: 0, options: VOLC_CLONE_LANGUAGES.map(value => ({ value })), visible: { condition: cloning } },
+        { id: 'volcCloneTranscript', type: 'textarea', order: 7, default: '', visible: { condition: cloning } },
+        { id: 'volcCloneDenoise', type: 'switch', order: 8, default: false, visible: { condition: cloning } },
+        { id: 'volcClonePreserveVolume', type: 'switch', order: 9, default: false, visible: { condition: cloning } },
+        { id: 'volcCloneExistingSlot', type: 'text', order: 10, default: '', visible: { condition: cloning } },
+        { id: 'volcCloneCustomId', type: 'text', order: 11, default: '', visible: { condition: () => false } },
     ],
-    endpoints: '/api/v3/tts/unidirectional',
+    runtimeConstraints: { mediaFields: [{ field: 'clone_audio', kind: 'audio' }] },
+    endpoints: { selector: params => cloning(params) ? '/api/v3/tts/voice_clone' : '/api/v3/tts/unidirectional' },
     request: {
         builder: (params) => {
-            const speaker = speechString(params.volcSeedIclSpeaker);
-            if (!speaker)
-                throw new Error('Seed-ICL 2.0 需要已创建的复刻音色 ID');
+            if (cloning(params)) {
+                const name = speechString(params.volcCloneName);
+                const source = Array.isArray(params.volcCloneAudio) ? params.volcCloneAudio[0] : params.volcCloneAudio;
+                const audio = typeof source === 'string' ? source : source && typeof source === 'object' && !Array.isArray(source) ? source.path ?? source.url : undefined;
+                const demo = speechText(params).trim();
+                if (!name)
+                    throw new Error('请填写音色名称');
+                if (typeof audio !== 'string' || !audio)
+                    throw new Error('请上传声音样本');
+                if (Array.from(demo).length < 4 || Array.from(demo).length > 300)
+                    throw new Error('克隆试听文本须为 4–300 字');
+                const language = params.volcCloneLanguage ?? 0;
+                if (!VOLC_CLONE_LANGUAGES.includes(language))
+                    throw new Error('请选择有效的录音语言');
+                const custom = speechString(params.volcCloneCustomId);
+                if (custom && !/^henji[a-zA-Z0-9]{8,64}$/.test(custom))
+                    throw new Error('克隆音色代号无效');
+                return {
+                    clone_audio: audio, voice_name: name, language,
+                    ...(speechString(params.volcCloneExistingSlot) ? { speaker_id: speechString(params.volcCloneExistingSlot) } : {}),
+                    ...(custom ? { custom_speaker_id: custom } : {}),
+                    ...(speechString(params.volcCloneTranscript) ? { text: speechString(params.volcCloneTranscript) } : {}),
+                    extra_params: { demo_text: demo, enable_audio_denoise: params.volcCloneDenoise === true, disable_volume_normalization: params.volcClonePreserveVolume === true },
+                };
+            }
+            const speaker = speechString(params.volcSeedIclSpeaker) ?? DEFAULT_SEED_TTS_20_VOICE;
+            const systemVoice = isSeedTts20SystemVoice(speaker);
+            if (!systemVoice && !/^S_/i.test(speaker) && params.volcIclActivationConsent !== true)
+                throw new Error('后付费音色首次正式合成将收取 138 元槽位费并锁定音色，请先确认费用');
             return {
-                resource_id: 'seed-icl-2.0',
-                req_params: { text: speechText(params), model: 'seed-tts-2.0-standard', speaker, audio_params: { format: 'mp3', sample_rate: 24000 } },
+                resource_id: systemVoice ? 'seed-tts-2.0' : 'seed-icl-2.0',
+                req_params: { text: speechText(params), ...(!systemVoice ? { model: 'seed-tts-2.0-standard' } : {}), speaker, audio_params: { format: 'mp3', sample_rate: 24000 } },
             };
         },
     },
-    pricing: { currency: '¥', calculator: () => Number.NaN, description: '火山语音按量计费单价未核实；不显示费用预估' },
+    pricing: { currency: '¥', calculator: params => speechCharacterPrice(params, 0.3), description: '3 元/万字符；后付费音色首次正式合成另收 138 元槽位费，训练试听按字符计费' },
 });
 export default volcengineSeedIcl20Model;

package/dist/catalog/volcengine-speech/seed-tts-2.0-voices.d.ts

@@ -0,0 +1,2 @@
+export declare const DEFAULT_SEED_TTS_20_VOICE = "zh_female_vv_uranus_bigtts";
+export declare const isSeedTts20SystemVoice: (id: string) => boolean;