{"info": {"title": "MiniMax Speech 2.8 HD异步语音合成", "version": "1.0.0", "description": "MiniMax 异步语音合成 API，支持多种音色、情绪、语速等参数设置，文本长度限制最长 5 万字符，支持文件输入（最长 10 万字符）"}, "paths": {"/v3/async/minimax-speech-2.8-hd": {"post": {"tags": ["Text to Audio"], "summary": "Text to Audio Async V2", "security": [{"BearerAuth": []}], "responses": {"200": {"content": {"application/json": {"schema": {"$ref": "#/components/schemas/T2AAsyncV2Resp"}, "examples": {"文件输入": {"value": {"file_id": 95157322514444, "task_id": "95157322514444", "base_resp": {"status_msg": "success", "status_code": 0}, "task_token": "eyJhbGciOiJSUz", "usage_characters": 101}}, "文本输入": {"value": {"file_id": 95157322514444, "task_id": "95157322514444", "base_resp": {"status_msg": "success", "status_code": 0}, "task_token": "eyJhbGciOiJSUz", "usage_characters": 101}}}}}, "description": "任务创建成功"}, "400": {"description": "请求参数错误"}, "401": {"description": "认证失败"}, "500": {"description": "服务器内部错误"}}, "description": "使用本接口，创建异步语音合成任务。支持文本或文件输入，文本长度限制最长 5 万字符，文件限制最长 10 万字符。", "operationId": "t2aAsyncV2", "requestBody": {"content": {"application/json": {"schema": {"$ref": "#/components/schemas/T2AAsyncV2Req"}, "examples": {"文件输入": {"value": {"text_file_id": 123456789, "voice_modify": {"pitch": 0, "timbre": 0, "intensity": 0, "sound_effects": "spacious_echo"}, "audio_setting": {"format": "mp3", "bitrate": 128000, "channel": 2, "audio_sample_rate": 32000}, "voice_setting": {"vol": 10, "pitch": 1, "speed": 1, "voice_id": "audiobook_male_1"}, "language_boost": "auto", "continuous_sound": false, "pronunciation_dict": {"tone": ["草地/(cao3)(di1)"]}}}, "文本输入": {"value": {"text": "真正的危险不是计算机开始像人一样思考(sighs)，而是人开始像计算机一样思考。计算机只是可以帮我们处理一些简单事务。", "voice_modify": {"pitch": 0, "timbre": 0, "intensity": 0, "sound_effects": "spacious_echo"}, "audio_setting": {"format": "mp3", "bitrate": 128000, "channel": 2, "audio_sample_rate": 32000}, "voice_setting": {"vol": 1, "pitch": 1, "speed": 1, "voice_id": "audiobook_male_1"}, "language_boost": "auto", "continuous_sound": false, "pronunciation_dict": {"tone": ["危险/dangerous"]}}}}}}, "required": true}}}}, "openapi": "3.0.3", "components": {"schemas": {"BaseResp": {"type": "object", "required": ["status_code", "status_msg"], "properties": {"status_msg": {"type": "string", "description": "状态详情"}, "status_code": {"type": "integer", "format": "int64", "description": "状态码

• `0`: 正常
• `1002`: 限流
• `1004`: 鉴权失败
• `1039`: 触发 TPM 限流
• `1042`: 非法字符超10%
• `2013`: 参数错误\n\n更多内容可查看 [错误码查询列表](/api-reference/errorcode) 了解详情"}}, "description": "本次请求的状态码及其详情"}, "VoiceModify": {"type": "object", "properties": {"pitch": {"type": "integer", "maximum": 100, "minimum": -100, "description": "音高调整（低沉/明亮），范围 [-100, 100]，数值接近 -100，声音更低沉；接近 100，声音更明亮"}, "timbre": {"type": "integer", "maximum": 100, "minimum": -100, "description": "音色调整（磁性/清脆），范围 [-100, 100]，数值接近 -100，声音更浑厚；数值接近 100，声音更清脆"}, "intensity": {"type": "integer", "maximum": 100, "minimum": -100, "description": "强度调整（力量感/柔和），范围 [-100, 100]，数值接近 -100，声音更刚劲；接近 100，声音更轻柔"}, "sound_effects": {"enum": ["spacious_echo", "auditorium_echo", "lofi_telephone", "robotic"], "type": "string", "description": "音效设置，单次仅能选择一种，可选值：\n1. spacious_echo（空旷回音）\n2. auditorium_echo（礼堂广播）\n3. lofi_telephone（电话失真）\n4. robotic（电音）"}}, "description": "声音效果器设置"}, "T2AAsyncV2Req": {"type": "object", "required": ["voice_setting"], "properties": {"text": {"type": "string", "description": "待合成音频的文本，限制最长 5 万字符。和 `text_file_id` 二选一必填

• 语气词标签：仅当模型选择 `speech-2.8-hd` 或 `speech-2.8-turbo` 时，支持在文本中插入语气词标签。支持的语气词：`(laughs)`（笑声）、`(chuckle)`（轻笑）、`(coughs)`（咳嗽）、`(clear-throat)`（清嗓子）、`(groans)`（呻吟）、`(breath)`（正常换气）、`(pant)`（喘气）、`(inhale)`（吸气）、`(exhale)`（呼气）、`(gasps)`（倒吸气）、`(sniffs)`（吸鼻子）、`(sighs)`（叹气）、`(snorts)`（喷鼻息）、`(burps)`（打嗝）、`(lip-smacking)`（咂嘴）、`(humming)`（哼唱）、`(hissing)`（嘶嘶声）、`(emm)`（嗯）、`(whistles)`（口哨）、`(sneezes)`（喷嚏）、`(crying)`（抽泣）、`(applause)`（鼓掌）"}, "text_file_id": {"type": "integer", "format": "int64", "description": "待合成音频的文本文件 id，单个文件长度限制小于 10 万字符，支持的文件格式：txt、zip。和 `text` 二选一必填，传入后自动校验格式。
• **txt 文件**：长度限制 <100000 字符。支持使用 `<#x#>` 标记自定义停顿。x 为停顿时长（单位：秒），范围 [0.01, 99.99]，最多保留两位小数。注意停顿需设置在两个可以语音发音的文本之间，不可连续使用多个停顿标记
• **zip 文件**：
• 压缩包内需包含同一格式的 txt 或 json 文件。
• json 文件格式：支持 [`title`, `content`, `extra`] 三个字段，分别表示标题、正文、附加信息。若三个字段都存在，则产出 3 组结果，共 9 个文件，统一存放在一个文件夹中。若某字段不存在或内容为空，则该字段不会生成对应结果"}, "voice_modify": {"$ref": "#/components/schemas/VoiceModify"}, "audio_setting": {"$ref": "#/components/schemas/T2AAsyncV2AudioSetting"}, "voice_setting": {"$ref": "#/components/schemas/T2AAsyncV2VoiceSetting"}, "aigc_watermark": {"type": "boolean", "default": false, "description": "控制在合成音频的末尾添加音频节奏标识，默认值为 False。该参数仅对非流式合成生效"}, "language_boost": {"enum": ["Chinese", "Chinese,Yue", "English", "Arabic", "Russian", "Spanish", "French", "Portuguese", "German", "Turkish", "Dutch", "Ukrainian", "Vietnamese", "Indonesian", "Japanese", "Italian", "Korean", "Thai", "Polish", "Romanian", "Greek", "Czech", "Finnish", "Hindi", "Bulgarian", "Danish", "Hebrew", "Malay", "Persian", "Slovak", "Swedish", "Croatian", "Filipino", "Hungarian", "Norwegian", "Slovenian", "Catalan", "Nynorsk", "Tamil", "Afrikaans", "auto"], "type": "string", "description": "是否增强对指定的小语种和方言的识别能力。默认值为 `null`，可设置为 `auto` 让模型自主判断。"}, "continuous_sound": {"type": "boolean", "default": false, "description": "启用该参数，使得子句衔接处更自然，仅支持 `speech-2.8-hd` 和 `speech-2.8-turbo` 模型"}, "pronunciation_dict": {"$ref": "#/components/schemas/T2AAsyncV2PronunciationDict"}}}, "T2AAsyncV2Resp": {"type": "object", "properties": {"file_id": {"type": "integer", "format": "int64", "description": "任务创建成功后返回的对应音频文件的 ID。

• 当任务完成后，可通过 file_id 调用 [文件检索接口](/api-reference/file-management-retrieve) 进行下载

• 当请求出错时，不返回该字段\n\n注意：返回的下载 URL 自生成起 9 小时（32400 秒）内有效，过期后文件将失效，生成的信息便会丢失，请注意下载信息的时间"}, "task_id": {"type": "string", "description": "当前任务的 ID"}, "base_resp": {"$ref": "#/components/schemas/BaseResp"}, "task_token": {"type": "string", "description": "完成当前任务使用的密钥信息"}, "usage_characters": {"type": "integer", "description": "计费字符数"}}}, "T2AAsyncV2AudioSetting": {"type": "object", "properties": {"format": {"enum": ["mp3", "pcm", "flac", "wav", "pcmu_raw", "pcmu_wav", "opus"], "type": "string", "default": "mp3", "description": "生成音频的格式。可选范围 [mp3, pcm, flac, wav, pcmu_raw, pcmu_wav, opus]，默认值为 mp3。pcmu_raw 与 pcmu_wav 为 G.711 μ-law 编码（采样率 8 kHz；pcmu_raw 为无文件头裸数据，pcmu_wav 封装在 WAV 容器中）。opus 为 Ogg/Opus 编码，仅支持采样率 [8000, 12000, 16000, 24000, 48000]，使用其他采样率会导致任务报错。"}, "bitrate": {"type": "integer", "format": "int64", "default": 128000, "description": "生成音频的比特率。可选范围 [32000, 64000, 128000, 256000]，默认值为 `128000`。该参数仅对 `mp3` 格式的音频生效"}, "channel": {"type": "integer", "format": "int64", "default": 2, "description": "生成音频的声道数。可选范围：[1, 2]，其中 `1` 为单声道，`2` 为双声道，默认值为 1"}, "audio_sample_rate": {"type": "integer", "format": "int64", "default": 32000, "description": "生成音频的采样率。可选范围 [8000, 16000, 22050, 24000, 32000, 44100]，默认为 `32000`"}}}, "T2AAsyncV2VoiceSetting": {"type": "object", "required": ["voice_id"], "properties": {"vol": {"type": "number", "format": "float", "default": 1, "maximum": 10, "minimum": 0, "description": "合成音频的音量，取值越大，音量越高。取值范围 (0, 10]，默认值为 1.0"}, "pitch": {"type": "integer", "default": 0, "maximum": 12, "minimum": -12, "description": "合成音频的语调，取值范围 [-12, 12]，默认值为 0，其中 0 为原音色输出"}, "speed": {"type": "number", "format": "float", "default": 1, "maximum": 2, "minimum": 0.5, "description": "合成音频的语速，取值越大，语速越快。取值范围 [0.5, 2]，默认值为1.0"}, "emotion": {"enum": ["happy", "sad", "angry", "fearful", "disgusted", "surprised", "calm", "fluent", "whisper"], "type": "string", "description": "控制合成语音的情绪，参数范围 [\"happy\", \"sad\", \"angry\", \"fearful\", \"disgusted\", \"surprised\", \"calm\", \"fluent\", \"whisper\"]，分别对应 8 种情绪：高兴，悲伤，愤怒，害怕，厌恶，惊讶，中性，生动，低语 \r
• 模型会根据输入文本自动匹配合适的情绪，一般无需手动指定 \r
• 该参数仅对 `speech-2.6-hd`, `speech-2.6-turbo`, `speech-02-hd`, `speech-02-turbo`, `speech-01-hd`, `speech-01-turbo` 模型生效 \r
• 选项 `fluent`, `whisper` 仅对 `speech-2.6-turbo`, `speech-2.6-hd` 模型生效"}, "voice_id": {"type": "string", "description": "合成音频的音色编号。若需要设置混合音色，请设置 timber_weights 参数，本参数设置为空值。支持系统音色、复刻音色以及文生音色三种类型，以下是部分最新的系统音色（ID），可查看 [系统音色列表](/faq/system-voice-id) 或使用 [查询可用音色 API](/api-reference/voice-management-get) 查询系统支持的全部音色\n
• **中文**:
• moss_audio_ce44fc67-7ce3-11f0-8de5-96e35d26fb85
• moss_audio_aaa1346a-7ce7-11f0-8e61-2e6e3c7ee85d
• Chinese (Mandarin)_Lyrical_Voice
• Chinese (Mandarin)_HK_Flight_Attendant
• **英文**:
• English_Graceful_Lady
• English_Insightful_Speaker
• English_radiant_girl
• English_Persuasive_Man
• moss_audio_6dc281eb-713c-11f0-a447-9613c873494c
• moss_audio_570551b1-735c-11f0-b236-0adeeecad052
• moss_audio_ad5baf92-735f-11f0-8263-fe5a2fe98ec8
• English_Lucky_Robot
• **日文**:
• Japanese_Whisper_Belle
• moss_audio_24875c4a-7be4-11f0-9359-4e72c55db738
• moss_audio_7f4ee608-78ea-11f0-bb73-1e2a4cfcd245
• moss_audio_c1a6a3ac-7be6-11f0-8e8e-36b92fbb4f95"}, "english_normalization": {"type": "boolean", "default": false, "description": "支持英语文本规范化，开启后可提升数字阅读场景的性能，但会略微增加延迟，默认 false"}}}, "T2AAsyncV2PronunciationDict": {"type": "object", "properties": {"tone": {"type": "array", "items": {"type": "string"}, "description": "定义需要特殊标注的文字或符号对应的注音或发音替换规则。在中文文本中，声调用数字表示：\n一声为 `1`，二声为 `2`，三声为 `3`，四声为 `4`，轻声为 `5`\n示例如下：\n[\"燕少飞/(yan4)(shao3)(fei1)\", \"omg/oh my god\"]"}}}}, "securitySchemes": {"BearerAuth": {"type": "http", "scheme": "bearer", "description": "`HTTP: Bearer Auth`\n - Security Scheme Type: http\n - HTTP Authorization Scheme: Bearer API_key，用于验证账户信息，可在 [账户管理 > 接口密钥](https://platform.minimaxi.com/user-center/basic-information/interface-key) 中查看。", "bearerFormat": "JWT"}}}}

speech-to-text