← Audio

MiniMax Audio（海螺语音）MiniMax Audio (Hailuo Voice)

MiniMax（稀宇科技）MiniMax (Xiyu Jizhi)

免费+订阅Free + subscription 中文TTSChinese TTS 声音克隆Voice cloning API向API-focused 国产直连Direct access in China partial lifecycle: active 核验Checked 2026-10-22

MiniMax Audio 是国产 TTS/声音克隆/API 三合一的音频线：中文语音质量口碑 + 音色克隆 + 音乐模型经 API 供给；网页与 API 两套入口两套条款。

MiniMax Audio is the domestic TTS/voice-clone/API three-in-one audio line: strong Chinese speech quality + voice cloning + music models served via API; web and API are two entries with two sets of terms.

适合与不适合When to use it, when not to

适合：中文 TTS、自有声音克隆工作流、API 集成、国产直连。不适合：克隆他人（授权红线）、开放权重、离线。

Fits: Chinese TTS, own-voice cloning workflows, API integration, domestic direct access. Not a fit: cloning others (consent red line), open weights, offline.

中文生态China availability 国产 / 国内可直接用 / 备案状态 · 快照domestic / direct access / CAC filing · snapshot 2026-10-22

出品属地Origin: 国产domestic　MiniMax（上海稀宇极智科技有限公司，中国）。MiniMax (Shanghai Xiyu Jizhi Technology Co., Ltd., China).
国内使用Direct access: 大陆可直连direct OK　minimax.io 大陆直连实测 200（2026-10-22）。minimax.io measured 200 direct from mainland (2026-10-22).
备案状态CAC filing: 已备案Filed：Shanghai-MiniMax-202505230046Shanghai-MiniMax-202505230046　·　MiniMax 模型备案号（百炼与百度智能云官方公示页交叉核验，2026-10-22）；主语是模型层——各产品入口的备案状态未逐一核验。MiniMax model-level filing number (cross-verified on Alibaba Cloud Bailian and Baidu Intelligent Cloud disclosure pages, 2026-10-22); the subject is the model layer — per-entry filing status was not verified one by one.

备案信息以网信办《生成式人工智能服务已备案信息》公示名单为准（按批次滚动更新，本站只抄编号不解读）；「大陆可直连」是本站从大陆网络直连实测的结果，不是厂商承诺。GitHub 类站点当日实测有波动，口径写在依据里。Filing status is quoted from the CAC public registry (updated batch by batch; this site copies the number and does not interpret it). "Direct access" is this site's own reachability measurement from a mainland network, not a vendor promise. GitHub-hosted sites fluctuated on the measurement day — the note says so.

固定坐标系Fixed coordinate system

8 个维度dimensions, 与同分区其他对象逐项可比each comparable item by item against others in the same section.

模型与开放条件Model & openness: MiniMax 自研语音/音频模型线（speech-02 系列等模型经 API 提供，官方文档叙事）；网页入口（hailuoai.com/audio 与 minimax.io/audio）+ API 平台（platform.minimaxi.com 线）双入口。
MiniMax's self-developed speech/audio model line (speech-02 series models served via API, official docs narrative); dual entry points: web (hailuoai.com/audio and minimax.io/audio) + API platform (platform.minimaxi.com line).
运行位置Where it runs: 网页 + API；生成在厂商云端。未见官方开放权重的音频模型声明（语言模型有开放权重先例，音频线未见）。
Web + API; generation on the vendor's cloud. No official open-weights audio-model statement seen (the language models have an open-weights precedent; the audio line does not).
本地文件Local files: 生成产物下载归本地；账户侧留存面未逐条核验。未见官方声明的本机数据面。
Generated outputs download to local; account-side retention surfaces not verified item by item. No officially stated local data surface seen.
关机后的任务Tasks after shutdown: 生成即时/任务制（API 调用即产即回）；无「关机续跑」形态声明—— 云端服务不依赖本机在线。
Generation is instant/task-based (API calls produce and return immediately); no "keep-running-after-shutdown" form stated — a cloud service does not depend on your machine being online.
工具与扩展Tools & extensions: TTS（多语言/多风格）、声音克隆/音色设计（官方产品面）、音乐生成线（音乐模型经 API 提供，官方文档叙事）；API 参数面（速度/音量/情绪控制）文档化。
TTS (multilingual/multi-style), voice cloning/voice design (official product surface), a music-generation line (music models served via API, official docs narrative); API parameter surfaces (speed/volume/emotion control) documented.
上下文与记忆Context & memory: 工具型产品无对话记忆概念；音色资产是它的「一致性」机制——克隆音色可复用不是记忆。
Tool-type product with no conversation-memory concept; voice assets are its "consistency" mechanism — cloned voices being reusable is not memory.
权限与限制Permissions & limits: 账号+订阅+API 计费体系（D9：价格不采）；声音克隆涉人格权——官方有授权口径页，细节未逐条核验（收录标准 3：克隆授权线必答）。备案：MiniMax 模型备案号 Shanghai-MiniMax-202505230046（百炼/百度云公示页交叉，2026-10-22）——模型层主语。
Account + subscription + API billing system (D9: pricing not collected); voice cloning involves personality rights — the official site has an authorization-framing page, details not verified item by line (inclusion criterion 3: the clone-authorization line is a must-answer). Filing: MiniMax model filing no. Shanghai-MiniMax-202505230046 (cross-checked on Bailian/Baidu Cloud disclosure pages, 2026-10-22) — the subject is the model layer.
适合什么任务Best for: 适合：中文 TTS 高质量向、声音克隆工作流（自有声音授权场景）、API 集成语音、国产直连。不适合：克隆他人声音（授权红线）、要开放权重、离线跑。
Fits: high-quality Chinese TTS, voice-cloning workflows (with authorization for your own voice), API-integrated voice, domestic direct connection. Does not fit: cloning other people's voices (authorization red line), wanting open weights, running offline.

头号误解Top misconception

声音克隆的「授权」是条款红线——克隆他人声音需可证明的授权，别按功能可用性理解边界Voice-clone 'consent' is a terms red line — cloning someone else's voice needs provable consent; don't read boundaries from feature availability
网页入口（hailuo/minimax.io）与 API 平台是两套配额与条款面——别互相推断The web entry (hailuo/minimax.io) and the API platform are two quota/terms surfaces — don't infer one from the other
备案号主语是 MiniMax 模型层——不等于每个产品入口单独备案The filing number's subject is the MiniMax model layer — it does not mean each product entry is separately filed

未知项清单Known unknowns

音频模型开放权重（未核验到官方声明）
各产品入口自身的备案状态（未逐一核验）
训练用途开关（未逐条核验）
音乐模型与语音模型的计费是否同池（未逐条核验）

Open weights for the audio models (no official statement verified)
Filing status of each product entry point itself (not verified one by one)
The training-use switch (not verified item by item)
Whether music and speech models share one billing pool (not verified item by item)

证据来源Evidence sources

判断可回到以下一手源复核。本站核验日Every judgement can be rechecked against the primary sources below. Verified on 2026-10-22，内容更新日content updated on 2026-10-22。

类型Type	名称Name	链接Link
official	MiniMax · 官网（产品与能力叙事）MiniMax · official site (product and capability narrative)	https://www.minimax.io
official	MiniMax Audio · 音频产品页（2026-10-22 直抓核验）MiniMax Audio · audio product page (direct-fetch verified 2026-10-22)	https://www.minimax.io/audio
changelog	MiniMax · News（官方更新线，changelog，2026-10-22 直抓核验）MiniMax · News (official update line, changelog, direct-fetch verified 2026-10-22)	https://www.minimax.io/news
docs	阿里云百炼 · 模型备案信息公示（备案号交叉核验源）Alibaba Cloud Bailian · model filing information publicity (filing-no. cross-verification source)	https://docs.bailian.console.aliyun.com/zh/model-studio/model-filing-information-publicity

实测记录Measured runs

本档未跑音频质量与延迟测试；结论只覆盖能力面与许可面。

实测建议（三条，都指向官方可读页没给的东西）：

克隆音色与官方音库的情绪表现差距
API 与网页入口的模型版本是否同步
声音克隆授权验证流程的实际严格度

This site has not completed measured runs. The coding-oriented test protocol is at protocols/coding/_protocol.md (the coding task suite does not apply to creation products).

Suggested measurements (three, all pointing at things the official readable pages do not give):

Emotional-performance gap between cloned voices and the official voice library
Whether API and web entry points run synchronized model versions
Actual strictness of the voice-clone authorization verification flow