Updated
What is Suno Speech?
Suno Speech turns words into expressive spoken audio. Send a script as text to suno-ai/speech and Suno reads it aloud, or send a short description and Suno writes the script first. With a script you can steer the delivery: tone describes the style in plain words, variety sets how much the read may vary, and background_music lays music under the voice or leaves it clean. vocal_gender picks the speaker in both modes. Each request returns two takes, so you can pick the one that sounds right. It suits narration, podcast intros, ads, explainers and voiceovers where a flat text-to-speech voice sounds too plain.
Key features of Suno Speech
Four features cover how Suno Speech works.
Your script or Suno's
Send text to have your exact words spoken, or description to have Suno write a script from a short brief and speak it.
Delivery in plain words
Describe the read with tone, such as "calm documentary narrator" or "energetic radio host", instead of tuning numeric voice settings.
Music on or off
Background music is on by default for a produced feel. Set background_music to false for a dry voice track you can mix yourself.
Two takes per request
Every call returns a pair of clips in the same shape as music generation. Listen to both and keep the better read.
What Suno Speech is best for
Podcast intros and outros
Give a script and a tone like "warm host, relaxed pace" and get a produced intro with music underneath.
Ads and promos
Describe the product and length and let Suno write and voice a short spot.
Explainers and narration
Read product walkthroughs or story narration with a consistent, natural delivery.
Video voiceovers
Turn background music off for a clean voice track to lay over your own edit.
Fast drafts
Try several tones on the same script before booking a voice actor.
Suno Speech use cases
Add a "Generate voiceover" button to a video editor that sends the scene script as text with background_music off, then drops the chosen take onto the timeline. Build an ad tool where a user types a product name and a length, sends it as description, and gets two ready-to-play spots back. For songs instead of speech, use Suno Music; for sound effects, use Suno Sound Effects.
Suno Speech API examples
All speech parameters are documented at docs.unifically.com/models/audio/suno/speech.
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "suno-ai/speech",
"input": {
"text": "Welcome back to the show. Today we are talking about the future of sound.",
"tone": "warm podcast host, relaxed pace",
"vocal_gender": "f",
"background_music": false
}
}'
The response contains a task_id. Poll GET /v1/tasks/<task_id> or set a callback_url on the request to receive the two speech clips when they are ready.
FAQs
People also ask about Suno Speech
Send text and Suno speaks exactly those words. Send description instead (for example "a 20 second radio ad for a coffee shop") and Suno writes the script for you, then speaks it. Each request takes one or the other, not both.
Yes. vocal_gender picks a female (f) or male (m) speaker. With a script, tone describes the delivery in plain words, such as "warm podcast host, relaxed pace", and variety (1 to 4) sets how much the delivery may vary from line to line.
By default, yes. With a script, set background_music to false for a clean voice-only track.
Two clips in clips, in the same shape as Suno music generation. Each clip has its own audio_url (an MP3) and clip ID, so you can compare both takes and keep the better one.
A flat rate per request, the same no matter how long the script is. Both returned takes are included.
