#When the ai voices will be able to actually sing (ANY voice)

1 messages · Page 1 of 1 (latest)

knotty cargoBOT
#
AI-generated Answer for witherking18

Short answer: ElevenLabs voices can sing today via Eleven Music, but Text-to-Speech (TTS) voices (including custom/cloned voices) do not have a documented “singing” mode, and ElevenLabs has not published an ETA for enabling singing with any arbitrary TTS voice.

Details:

  • Eleven Music generates full songs with vocals and lyrics from prompts in multiple languages; it’s available in the web app now. The docs explicitly list “Vocals & Lyrics” as a feature. (help.elevenlabs.io, elevenlabs.io)
  • A prompting guide shows how to request a cappella vocals and control lyric timing, confirming singing support within Eleven Music. (elevenlabs.io)
  • As of August 20, 2025, a public Music Generation API is available to paid users (compose/stream endpoints). (elevenlabs.io)
  • The Speech Synthesis API documentation (TTS) does not document singing controls or melody input, and there’s no announcement in the docs/changelog that arbitrary TTS voices can sing. (elevenlabs.io)

What you can do today:

  • Use Eleven Music to create sung vocals, including a cappella, and then mix them with other audio. (elevenlabs.io)

If you meant “when will my specific cloned TTS voice be able to sing,” there’s no published timeline in the docs. Keep an eye on the product changelog for any updates. (elevenlabs.io)

Sources:
What is Eleven Music? – ElevenLabs
Models | ElevenLabs Documentation
Prompting Eleven Music | ElevenLabs Documentation
August 20, 2025 | ElevenLabs Documentation
Create speech | ElevenLabs Documentation
Prompting Eleven Music | ElevenLabs Documentation
Changelog | ElevenLabs Documentation

regal creek
# knotty cargo

I still want to know when there will be a thing that will allow the ai voices to sing

knotty cargoBOT
#
AI-generated Answer for witherking18

Short answer: it’s available now via Eleven Music.

  • Eleven Music generates full songs from text prompts, including AI vocals and lyrics in multiple languages. It launched in the web app on August 5, 2025. (elevenlabs.io, help.elevenlabs.io)
  • The Eleven Music API for developers became available on August 18, 2025. (elevenlabs.io)

Important note: Eleven’s public docs describe singing as part of the Eleven Music product. They do not document a way to make existing TTS voices or cloned voices “sing” through the TTS endpoints; if you need singing today, use Eleven Music. (elevenlabs.io, help.elevenlabs.io)

If you want, I can point you to the Music prompting guide and usage terms next. (help.elevenlabs.io)

Sources:
Eleven Music is Here | ElevenLabs
What is Eleven Music? – ElevenLabs
Eleven Music, now available in the API | ElevenLabs
Eleven Music is Here | ElevenLabs
What is Eleven Music? – ElevenLabs
What is Eleven Music? – ElevenLabs

regal creek
# knotty cargo

Dude, I don't want the music I want the singing! singing voices! that's what I want! I want my ai voices that I create to be able to sing, I have a song to publish!

knotty cargoBOT
#
AI-generated Answer for witherking18

Short answer: ElevenLabs supports AI singing today via Eleven Music, but there’s no documented way to make your own cloned/created “My Voices” sing yet.

What you can do now

  • Generate full songs with AI vocals and lyrics (multi‑language), edit sections/lyrics, and export for production using Eleven Music on the website or via the new Music API. (elevenlabs.io, help.elevenlabs.io)
  • Use the results commercially if you’re on a paid plan, subject to the Music Terms (there are plan‑specific restrictions for contexts like TV/film, ads, and some enterprise uses). Review the Music Terms before publishing/distribution. (help.elevenlabs.io, elevenlabs.io)

What isn’t documented

  • Eleven’s Text‑to‑Speech/Voice Changer features are documented for speech, not singing, and Eleven Music docs/blog do not describe selecting a custom “My Voices” clone/design for singing vocals. There’s no published ETA for that capability. (elevenlabs.io, help.elevenlabs.io)

How to move forward for your release

  • Use Eleven Music to create the sung vocals/track, then export MP3. Confirm your intended distribution fits your plan’s Music Terms before publishing (e.g., ads/film/TV may require Enterprise/approval). (help.elevenlabs.io, elevenlabs.io)

If you need custom‑voice singing specifically, there isn’t a supported, documented workflow today. I’ll share an update when Eleven publishes one. (help.elevenlabs.io)

Sources:
Eleven Music is Here | ElevenLabs
What can I generate with Eleven Music? – ElevenLabs
Can I publish the content I generate on the platform? – ElevenLabs
Music Terms | ElevenLabs
Text to Speech | ElevenLabs Documentation
What is Eleven Music? – ElevenLabs
What can I generate with Eleven Music? – ElevenLabs
Music Terms | ElevenLabs
What is Eleven Music? – ElevenLabs

regal creek
knotty cargoBOT
#
AI-generated Answer for witherking18

I hear you. Here’s the straight, official answer:

  • Yes—voices can sing today via Text-to-Speech with Eleven v3 (Alpha). Add the [sings] audio tag around your lyrics (e.g., “[sings] …lyrics…”). This tag is experimental: results vary by voice and aren’t guaranteed for every voice or passage. For best responsiveness, use v3 with Creative/Natural stability. (elevenlabs.io, help.elevenlabs.io)

  • If you need full songs (instrumental + AI vocals + AI lyrics), use Eleven Music. It generates complete tracks from text prompts and supports multilingual vocals. It’s available in the web app now, and as of August 18, 2025, also in the public API. (help.elevenlabs.io, elevenlabs.io)

  • Can “any” custom voice you create be used to sing? With TTS v3, you can select your cloned/designed voice and try the [sings] tag—but because the tag is experimental, quality and reliability will differ by voice. For Eleven Music specifically, the docs do not state that you can choose a specific “My Voices” voice as the vocalist; I can’t find an official way to do that today. (elevenlabs.io, help.elevenlabs.io)

  • Planning to publish commercially? Eleven Music is cleared for broad commercial use (see Music Terms on the product/help pages). For TTS outputs, you own the audio and can use it commercially on paid plans. (help.elevenlabs.io) (elevenlabs.io)

Quick how-to for singing with your voice (TTS v3):

  1. In Speech Synthesis, pick model “Eleven v3 (Alpha)” and select your voice. 2) Enter your lyrics and wrap them with “[sings] …”. 3) If it doesn’t sing, try a different voice or increase creativity (stability) as suggested in the v3 prompting guide. (help.elevenlabs.io, elevenlabs.io)

If you want full-song generation instead, open Eleven Music, describe the genre/mood/sections, and include that you want vocals and the language; iterate section-by-section for control. API is available if you’re building this into a workflow. (elevenlabs.io, help.elevenlabs.io)

If you need something more specific (e.g., “make this exact cloned voice sing this chorus on this melody”), there’s no official, documented feature for melody control or voice selection in Eleven Music yet. I’ll gladly point you to it when ElevenLabs publishes it. (help.elevenlabs.io)

regal creek
#

Okay, but how am I going to do it, can you give me a detailed toturial?

knotty cargoBOT
#
AI-generated Answer for witherking18

Short answer: today, the supported way to get ElevenLabs to sing is Eleven Music (text‑to‑song with vocals and lyrics). The Speech Synthesis/TTS endpoints and custom voices in My Voices are for spoken audio; there’s no documented parameter to make a specific TTS voice “sing.” Use Eleven Music instead.

Step‑by‑step (no code, in the browser)

  1. Open Eleven Music and create a new track. In your first prompt, be explicit about genre, tempo, mood, language, and that you want vocals (e.g., “modern pop ballad, 90 BPM, female vocals in English”). Generate.
  2. Edit the composition section‑by‑section. Use the timeline to add/delete sections (Verse/Chorus/Bridge), click a section’s text box to type or replace lyrics, and use Include/Exclude Styles for finer control. Click Generate to apply changes.
  3. Iterate with better prompts. The help center’s prompting tips show how to get reliable vocals and structure by being specific and building the song piece‑by‑piece.
  4. Export your song. Eleven Music offers studio‑ready MP3 export from the UI.

Step‑by‑step (API)

  1. You’ll need a paid plan to use the Eleven Music API. Create an API key.
  2. Compose music from a prompt via the SDK/API (e.g., music.compose with a prompt and music_length_ms). For deeper control, generate a composition_plan, then compose from that plan; the detailed endpoint returns the plan and any lyrics in the result for inspection.
  3. API availability note: the Music API and Eleven v3 API were made broadly available on August 20, 2025; see the changelog.
  4. The product page also shows a minimal compose example and links to docs.

Commercial use

  • Eleven Music is cleared for broad commercial use; review the Music Usage Terms linked from the help article before publishing.

If you specifically need a cloned voice to sing: there isn’t a documented way to make a My Voices/PVC voice perform pitched singing via the TTS endpoints today; use Eleven Music for sung vocals.

Want me to walk you through building a first verse/chorus prompt and section edits for your song?

regal creek
#

I know how to edit songs or stuff like that. I want to know how. do. I. make. a. voice. I. designed. SING IN THE RIGHT WAY.