Text-to-Speech (TTS)

Text-to-Speech converts text input into audio output. Otoroshi exposes TTS through an OpenAI-compatible API endpoint.

Supported providers and models

Provider	Models	Default voice
OpenAI	tts-1, tts-1-hd, gpt-4o-mini-tts	alloy
Azure OpenAI	tts, tts-hd, gpt-4o-mini-tts	alloy
Cloud Temple 🇫🇷 🇪🇺	tts-1, tts-1-hd, gpt-4o-mini-tts	alloy
Groq	playai-tts, playai-tts-arabic	alloy
ElevenLabs	eleven_monolingual_v1, eleven_multilingual_v2	Rachel

Available voices

OpenAI / Azure OpenAI / Cloud Temple

alloy, ash, ballad, coral, echo, fable, onyx, nova, sage, shimmer, verse

Groq

Arista-PlayAI, Atlas-PlayAI, Basil-PlayAI

ElevenLabs

Voices are fetched dynamically from the ElevenLabs API. You can browse available voices in the Otoroshi admin UI when configuring the audio model entity.

TTS configuration

OpenAI / Azure OpenAI / Cloud Temple

{
  "tts": {
    "enabled": true,
    "model": "gpt-4o-mini-tts",
    "voice": "alloy",
    "instructions": "Speak in a friendly tone",
    "response_format": "mp3",
    "speed": 1.0
  }
}

Parameter	Type	Default	Description
enabled	boolean	true	Enable or disable TTS
model	string	gpt-4o-mini-tts	The TTS model to use
voice	string	alloy	The voice to use for generation
instructions	string	—	Optional instructions for the voice (e.g. tone, style)
response_format	string	—	Audio format: mp3, opus, aac, flac, wav, pcm
speed	number	—	Speed of the generated audio (0.25 to 4.0)

Groq

{
  "tts": {
    "enabled": true,
    "model": "playai-tts",
    "voice": "Arista-PlayAI",
    "response_format": "wav"
  }
}

Same parameters as OpenAI.

ElevenLabs

{
  "tts": {
    "enabled": true,
    "model_id": "eleven_monolingual_v1",
    "voice_id": "21m00Tcm4TlvDq8ikWAM",
    "output_format": "mp3_44100_128"
  }
}

Parameter	Type	Default	Description
enabled	boolean	true	Enable or disable TTS
model_id	string	eleven_monolingual_v1	The ElevenLabs model ID
voice_id	string	21m00Tcm4TlvDq8ikWAM	The ElevenLabs voice ID (Rachel by default)
output_format	string	mp3_44100_128	Output audio format

API usage

Plugin setup

Add the Cloud APIM - Text to speech backend plugin to your route:

{
  "enabled": true,
  "plugin": "cp:otoroshi_plugins.com.cloud.apim.otoroshi.extensions.aigateway.plugins.OpenAICompatTextToSpeech",
  "config": {
    "refs": ["audio-gen-model_xxxxxxxxx"]
  }
}

Request

curl https://my-audio-endpoint.example.com/v1/audio/speech \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $OTOROSHI_API_KEY" \
  -d '{
    "input": "Hello, how are you today?",
    "model": "gpt-4o-mini-tts",
    "voice": "alloy",
    "response_format": "mp3",
    "speed": 1.0
  }' \
  --output speech.mp3

The response is a streamed audio file with the appropriate content type.

Request parameters

Parameter	Type	Description
input	string	The text to convert to speech (required)
model	string	Model to use. Supports provider_id###model_name routing syntax
voice	string	Voice to use
instructions	string	Optional voice instructions
response_format	string	Desired audio format
speed	number	Playback speed multiplier

Full entity example with OpenAI

{
  "id": "audio-gen-model_xxxxxxxxx",
  "name": "OpenAI Audio",
  "description": "OpenAI text-to-speech",
  "provider": "openai",
  "config": {
    "connection": {
      "token": "${vault://local/OPENAI_API_TOKEN}",
      "timeout": 30000
    },
    "options": {
      "tts": {
        "enabled": true,
        "model": "gpt-4o-mini-tts",
        "voice": "alloy",
        "response_format": "mp3",
        "speed": 1
      },
      "stt": {
        "enabled": true,
        "model": "whisper-1"
      },
      "translation": {
        "enabled": true,
        "model": "whisper-1"
      }
    }
  },
  "kind": "ai-gateway.extensions.cloud-apim.com/AudioModel"
}

Full entity example with Groq

{
  "id": "audio-gen-model_xxxxxxxxx",
  "name": "Groq TTS",
  "description": "Groq text-to-speech",
  "provider": "groq",
  "config": {
    "connection": {
      "token": "${vault://local/GROQ_API_TOKEN}",
      "timeout": 30000
    },
    "options": {
      "tts": {
        "enabled": true,
        "model": "playai-tts",
        "response_format": "wav"
      },
      "stt": {
        "enabled": true,
        "model": "whisper-large-v3-turbo"
      },
      "translation": {
        "enabled": true,
        "model": "whisper-large-v3"
      }
    }
  },
  "kind": "ai-gateway.extensions.cloud-apim.com/AudioModel"
}