Chatterbox Text to Speech API — English, WAV

Name: Chatterbox API
SKU: resemble-ai/chatterbox-v1/text-to-speech
Price: 0.06 USD
Availability: InStock

Introduction

What it returns

One WAV file per task: 24 kHz, 16-bit, mono PCM. There are no word timings or subtitles in the result.

The voice

Chatterbox is Resemble AI's open-source English model (MIT licence). Without a reference it reads in its built-in voice; there is no voice list. Send a short MP3 or WAV clip in reference_audio_url to clone a voice: the reading takes on that clip's timbre, not its words. A few seconds of one clean voice without music works best, and you must have the right to use the voice.

Controls

exaggeration (0–1, default 0.5): how emotional the reading is.
temperature (0.05–2, default 0.7): variation between takes.
guidance_scale (0.1–1, default 0.5): how closely the reading follows the text's pacing; around 0.3 slows a fast, dramatic delivery.

Tags such as <laugh>, <sigh> or [gasp] are read aloud as words in our tests. Spell sounds out instead: "ha…", "mmm…", "ahh…".

Length

text accepts up to 1,000 characters. A 270-character paragraph came back as about 15 seconds of audio in our tests; most tasks finished in 10 to 30 seconds.

Request parameters

Parameter	Required	Description
`exaggeration`	No	How emotional the reading is, from 0 (flat) to 1 (dramatic). ≥ 0 · ≤ 1 · Default: 0.5
`guidance_scale`	No	How strictly the model follows the prompt. ≥ 0.1 · ≤ 1 · Default: 0.5
`reference_audio_url`	No	Optional clip of the voice to imitate; the reading takes on its timbre.
`temperature`	No	Variation between takes, from 0.05 to 2. ≥ 0.05 · ≤ 2 · Default: 0.7
`text`	Yes	Up to 1,000 characters of English, read exactly as written. Spell sounds out; tags are read aloud.

Pricing

Prices are in credits, the same credits you use on the website (1 credit = $0.01). The hold is taken when the task is accepted and settled to the actual charge when it finishes; failed tasks are refunded.

Option	Credits per run	≈ USD
Standard	5.4	$0.054

Each task is rounded up to whole credits, so one run at the lowest tier costs 6 credits per run ($0.06).

For an exact number before you run, send the same body to POST /jobs/quote.

API quick start

Call the Chatterbox API in three steps

Create an API key. Sign in to PopNow and create a key in the API console. Keys start with sk-popnow- and use the same credits as the website.
Send the request. POST /jobs/createTask with model resemble-ai/chatterbox-v1/text-to-speech and an input object that follows the parameter table. The response returns a task ID straight away.
Get the result. Poll GET /jobs/recordInfo with the task ID, or pass a callBackUrl and we post the result to it. Download the output links when the task succeeds.

# 1. Create the task. Keep the Idempotency-Key if you retry after a network error.
export IDEMPOTENCY_KEY="$(uuidgen)"
curl -X POST https://api.popnow.ai/api/v1/jobs/createTask \
  -H "Authorization: Bearer $POPNOW_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $IDEMPOTENCY_KEY" \
  -d '{
  "model": "resemble-ai/chatterbox-v1/text-to-speech",
  "input": {
    "text": "Welcome back. Your files finished rendering a minute ago and are ready to download."
  }
}'

# 2. Copy data.taskId from the response, then poll until state is succeeded, failed, canceled or expired.
TASK_ID="TASK_ID_FROM_RESPONSE"
curl "https://api.popnow.ai/api/v1/jobs/recordInfo?taskId=$TASK_ID" \
  -H "Authorization: Bearer $POPNOW_API_KEY"

Replace the prompt and any links with your own. The API tab has the response, polling and callback reference.

Chatterbox API questions

How much does the Chatterbox API cost?

Chatterbox costs 6 credits ($0.06) per run. Credits are the same as on the website (1 credit = $0.01), new accounts get 30 free credits, and failed tasks are refunded.

What is the model name for Chatterbox in the API?

Put resemble-ai/chatterbox-v1/text-to-speech for Text to Speech in the model field of POST /jobs/createTask.

Can I choose a voice?

There is no voice list, but you can clone one: send a short MP3 or WAV clip in reference_audio_url and the reading takes on that voice's timbre. Without a clip, every request uses Chatterbox's built-in voice. The sliders change how the voice performs, not who it is. Use only voices you have the right to use.

Do tags such as <laugh> or [sigh] work?

No. In our tests the model read tags like these aloud as words. Spell sounds out instead — "ha…", "mmm…", "ahh…" — and raise exaggeration for a more emotional reading.