🎙️

Text-to-Speech & Voice Cloning

Natural speech in 23 languages, with optional voice cloning.

Send up to 400 characters of text, get a WAV back. Add a 5-20 second reference clip to speak in that voice. Powered by Chatterbox (MIT licence).

🚀 Open Text-to-Speech API reference

What it does

One endpoint for standard narration and zero-shot voice cloning.

23 languages

Arabic, Danish, German, Greek, English, Spanish, Finnish, French, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Dutch, Norwegian, Polish, Portuguese, Russian, Swedish, Swahili, Turkish and Chinese.

Voice cloning

Upload a clean 5-20 second clip of the speaker as voice_ref. You must have the speaker's consent. Same price as standard speech.

Engines

chatterbox-turbo for English, chatterbox-multilingual for the other languages. auto (default) picks for you.

Output

WAV audio. Up to 400 characters per request; split longer scripts into several requests.

API

Authenticate with Authorization: Bearer YOUR_API_KEY (create a key in the dashboard). Submit a multipart form, then poll the status endpoint until status is completed and read output_url.

POST /v1/tts/generate

Returns an id. Poll GET /v1/tts/status/{id} for the WAV.

FieldTypeRequiredDescription
textstringyesText to speak, max 400 characters.
languagestringnoDefault en. One of: ar, da, de, el, en, es, fi, fr, he, hi, it, ja, ko, ms, nl, no, pl, pt, ru, sv, sw, tr, zh.
voice_reffilenoReference voice clip for cloning (WAV/MP3/M4A, 5-20 s of clean speech, max 10 MB).
modelstringnoauto (default), chatterbox-turbo or chatterbox-multilingual.
curl -X POST https://api.pixelapi.dev/v1/tts/generate \ -H "Authorization: Bearer YOUR_API_KEY" \ -F "text=Welcome to our store. Today's deals end at midnight." \ -F "language=en" \ -F "voice_ref=@my_voice_10s.wav" # then poll until status is "completed" curl https://api.pixelapi.dev/v1/tts/status/TTS_JOB_ID -H "Authorization: Bearer YOUR_API_KEY"

Pricing

Billed in credits. 1 credit = $0.001 on the Starter plan; the cash value of a credit depends on your plan. Failed jobs are refunded.

Speech
40 credits / 1,000 chars
Charged per character, minimum 1 credit. ≈ $0.04 per 1,000 characters on Starter.
Voice cloning
Same price
Adding a voice_ref clip does not change the price.

FAQ

How much does text-to-speech cost?

40 credits per 1,000 characters, charged per character with a minimum of 1 credit. Voice cloning costs the same. 1 credit = $0.001 on the Starter plan.

How long can the text be?

Up to 400 characters per request. Split longer scripts into several requests and join the WAV files.

Which languages are supported?

Arabic, Danish, German, Greek, English, Spanish, Finnish, French, Hebrew, Hindi, Italian, Japanese, Korean, Malay, Dutch, Norwegian, Polish, Portuguese, Russian, Swedish, Swahili, Turkish and Chinese.

How does voice cloning work?

Attach a clean 5-20 second recording of the speaker as voice_ref. Only clone voices you have permission to use.

Which model is used?

Chatterbox (MIT licence): chatterbox-turbo for English and chatterbox-multilingual for other languages.

Ready to try Text-to-Speech?

New accounts get 300 free trial credits (one trial per person), no card required.

🚀 Open Text-to-Speech ← All tools