Audio API

AI Audio & Speech API

Text to speech, voice cloning, transcription, and sound effects with ElevenLabs, Fish Audio, Google and the public audio catalog.

32 audio models · public audio endpoints
32
Audio models
6
Providers
89
Public endpoints
Playground live
Model
Text
Your next audio render starts here. Choose a public model to inspect its current price and endpoint details.
Voice Default voice
Format mp3
Speed 1.0x
Mode text_to_speech

Examples

Hear the Voices

Aria — Natural

"Every model, every provider, one API key. That's what RunAPI gives you."

Miles — Narration

"Generate video, music, images, and audio from a single integration."

Nova — Conversational

"Switch between providers without changing a single line of code."

Kai — Broadcast

"Production-ready AI infrastructure with built-in failover and monitoring."

Everything the audio endpoint can do

Text to Speech

Stream natural speech from text with per-request voice and speed control.

Voice Cloning

Build a reusable voice from a short reference sample, then call it by id like any other voice.

Speech to Text

Transcribe long-form audio with timestamps, speaker labels, and word-level confidence.

Text to Sound

Generate foley and sound effects from a description — impacts, ambience, UI cues.

Text to Dialogue

Render multi-speaker conversations with distinct voices and natural turn timing in one call.

Audio models on one key

ElevenLabs/audio-isolation

ElevenLabs

$0.12 / minute

ElevenLabs/sound-effect-v2

ElevenLabs

$0.15 / minute

ElevenLabs/speech-to-text

ElevenLabs

$0.04 / minute

ElevenLabs/text-to-dialogue-v3

ElevenLabs

$0.14 / 1K chars

ElevenLabs/text-to-speech-multilingual-v2

ElevenLabs

$0.12 / 1K chars

ElevenLabs/text-to-speech-turbo-v2.5

ElevenLabs

$0.06 / 1K chars

View all 32 audio models
Quickstart

Synthesize speech

curl https://runapi.ai/api/v1/elevenlabs/text_to_speech \
  -H "Authorization: Bearer $RUNAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"text-to-speech-multilingual-v2","text":"Welcome to RunAPI.","voice":"Adam","language_code":"en"}'
How it works

Four Steps to Your First Generation

  1. Get Your API Key

    One key for all TTS and voice models.

  2. Pick a Voice

    Browse available voices across supported audio models.

  3. Send Text

    POST your text and voice selection to /api/v1/elevenlabs/text_to_speech.

  4. Get Audio

    Receive the generated audio from the task result.

Why Developers Choose RunAPI

One Key, All Models

Stop juggling API keys. One integration gives you access to 230+ models across all providers.

15-25% Cost Savings

Our volume agreements with providers mean lower prices than going direct. No markup on top.

Automatic Failover

If a provider goes down, we route to the next best option. Your app stays up.

OpenAI-Compatible

Drop-in replacement for OpenAI SDK. Change one line of code to access any model.

Real-Time Billing

Pay only for what you use. Per-request pricing with no minimums. Track costs per model, per project.

API Key Authentication

Authenticate audio API requests with your RunAPI API key.

What teams build with it

Voice Agents

Pair streaming TTS with an LLM and transcription for a full duplex voice loop on one key.

Localization

Dub existing video into new languages with a cloned voice that stays recognisable.

Meeting Notes

Transcribe with speaker labels, then summarise through the LLM endpoint in the same pipeline.

Accessibility

Add narration to articles, docs, and in-app content without a recording session.

RunAPI vs alternatives

Feature RunAPI ElevenLabs Direct OpenAI TTS
Voice models 32 audio models 1 provider 6 voices
Pricing From $0.12 / minute Official rate Official rate
Real-time streaming Yes Yes No
Unified billing Yes No No
MCP server Yes No No
Provider failover Yes No No

Audio API questions

How is TTS billed?

Per 1,000 characters of input text at the model's rate. Transcription is billed per minute of audio.

Can I clone a voice?

Yes. Upload a reference sample to create a voice id, then pass that id on any later request.

Is streaming supported?

Streaming support depends on the selected endpoint. Check its documented response format before integrating playback.

Which languages are covered?

Language support varies by model. Check the selected model's documentation for supported languages.

Do you keep audio or voice samples?

Request and response content may be retained. See the Privacy Policy for data handling details.

Can I get word-level timestamps?

Timestamp support and granularity depend on the selected transcription endpoint. Check its documented response fields; speaker labels are a separate capability.

Give your product a voice

Speech, cloning, and transcription behind one key. Start free, no credit card required.

Sign in to generate

Create an API key to run audio requests through your account.