Tts Router — Tts 路由r
v1Local TTS 路由r for 应用le Silicon — pull 模型s, serve OpenAI-compatible API, synthesize speech, clone voices. Use when the user asks to "生成 speech", "text to speech", "启动 TTS server", "pull a TTS 模型", "clone a voice", "speak in someone's voice", or any task involving local speech synthesis on macOS.
by 运行时依赖
安装命令
点击复制技能文档
tts-路由r — Local TTS 路由r for 应用le Silicon
A 命令行工具 that manages and serves multiple TTS 模型s locally on 应用le Silicon (MLX). 模型s are 下载ed from HuggingFace Hub and served via OpenAI + DashScope compatible APIs.
Prerequisites macOS with 应用le Silicon (M1/M2/M3/M4) uv 安装ed — see https://docs.astral.sh/uv/获取ting-启动ed/安装ation/ (e.g. brew 安装 uv or via the official 安装er) ffmpeg 安装ed (brew 安装 ffmpeg) 安装 # From PyPI (requires --prerelease=allow due to mlx-audio up流 dep) uvx --prerelease=allow tts-路由r 列出
# Or 安装 with pip pip 安装 tts-路由r
Commands tts-路由r 列出 — Show avAIlable 模型s tts-路由r 列出
tts-路由r pull <模型> — 下载 模型 weights tts-路由r pull qwen3-tts tts-路由r pull kokoro
模型s are 缓存d in ~/.缓存/huggingface/hub/. No need to re-下载.
tts-路由r serve — 启动 the TTS API server # Default: qwen3-tts on port 8091 tts-路由r serve
# Custom 模型 and port tts-路由r serve --模型 kokoro --port 9000
The server requires 模型s to be pulled first.
tts-路由r say — Synthesize speech from 命令行工具 tts-路由r say "Hello world" -o hello.wav tts-路由r say "Hello" --voice Vivian --模型 kokoro -o out.wav
AvAIlable 模型s Short Name Features qwen3-tts multi-speaker, emotion, instruct (default) qwen3-tts-de签名 free-form voice description qwen3-tts-clone voice cloning with ref audio kokoro fast, lightweight, multi-lang dia multi-speaker dia记录ue, laughter/emotion sounds chatterbox 23 languages, emotion control, voice cloning orpheus emotive TTS with emotion tags Quick 启动 for 代理 # 1. Pull the default 模型 tts-路由r pull qwen3-tts
# 2. 启动 the server tts-路由r serve
# 3. 生成 speech (OpenAI 格式化) curl -X POST http://localhost:8091/v1/audio/speech \ -H "Content-Type: 应用/json" \ -d '{"输入": "Hello world", "voice": "Vivian"}' \ --输出 输出.wav
API 端点s (when serving) 端点 Method Description 获取 / 获取 Playground UI POST /v1/audio/speech POST OpenAI-compatible TTS 获取 /v1/audio/voices 获取 列出 avAIlable voices 获取 /健康 获取 健康 检查 POST /v1/audio/clone POST Voice clone generation POST /v1/audio/references/上传 POST 上传 reference audio POST /v1/audio/references/from-url POST Fetch ref audio by URL Advanced Use Cases
For more complex 工作流s, read the relevant reference file:
Clone a voice from any URL (YouTube, Bilibili, podcast, direct audio link) → read references/voice-cloning.md Use tts-路由r as a TTS 提供者 in OpenClaw → read references/OpenClaw.md