AgentSkillsCN

openclaw-groq-orpheus-tts

为OpenClaw提供安全的配置变更，并支持自动回滚。在打补丁前进行备份，重启后执行健康检查，若出现故障则自动回滚。命令包括：patch、restore、list、diff、validate、doctor。

SKILL.md
--- frontmatter
name: openclaw-groq-orpheus-tts
description: Fast and FREE Arabic (Saudi) & English AI voices. Generous 100 requests per day via Groq API. Professional, high-quality generative TTS using Orpheus models.
metadata: {"openclaw":{"emoji":"🎙️","requires":{"bins":["curl","ffmpeg"],"env":["GROQ_API_KEY"]},"primaryEnv":"GROQ_API_KEY"}}

Groq Orpheus TTS

A powerful and fast text-to-speech skill that leverages Groq's Orpheus models.

Key Features:

  • •100% Free Tier Friendly: Uses the Groq Free API key (100 requests per day).
  • •Ultra-Fast: Near-instant audio generation.
  • •High Quality: Professional generative AI voices.

Supported Languages:

  • •Arabic: Authentic Saudi dialect synthesis (Voices: fahad, sultan, noura, lulwa, aisha).
  • •English: Expressive, high-quality speech (Voices: autumn, diana, hannah, austin, daniel, troy).

Requirements

  • •API Key: A GROQ_API_KEY from the Groq Console. This is FREE.
  • •Terms: You must accept the model terms for orpheus-v1-english on the Groq Playground before using the English model.
  • •Tools: ffmpeg must be installed on your system for audio conversion.

Requirements

  • •API Key: GROQ_API_KEY from Groq Console.
  • •Terms: You must accept the model terms for orpheus-v1-english on the Groq Playground before using the English model.
  • •Tools: ffmpeg must be installed on your system for audio conversion.

Usage

You can ask the assistant to say something or generate an audio file.

Voices Available

  • •Arabic (ar): fahad (Male), sultan (Male), noura (Female), lulwa (Female), aisha (Female).
  • •English (en): autumn, diana, hannah, austin, daniel, troy.

Commands

bash
# General usage
./groq-tts.sh "Text" output.mp3 [voice] [lang]

# Examples
./groq-tts.sh "أهلا بك" welcome.mp3 fahad ar
./groq-tts.sh "Hello world" hello.mp3 troy en

Chat Responses

When you want the assistant to reply in voice, use:

bash
./groq-tts.sh "Your message" /tmp/reply.mp3 fahad ar
# Then include MEDIA:/tmp/reply.mp3 in the response.