AI MUSIC, ANSWERED

Can Suno, Udio or Lyria Translate a Song?

FAMILIAR LABS RESEARCHALL ARTICLES

ABSTRACT

Not the song you know. Suno, Udio, Google's Lyria and the other AI music generators make a new song from a prompt: give one lyrics in another language and a new vocal sings them. The artist's own recording does not cross over. Song Dubbing from Familiar Labs carries that recording, in the artist's voice, into another language. Nothing is sung again.
SUNOUDIOLYRIAELEVENMUSIC

Can Suno translate a song?

Suno's closest feature is Covers, launched in September 2024: upload a track or pick one from your library, and Suno remakes it. In Suno's words, Covers works by:

keeping the melody and adapting the track to a different style
Suno, Introducing Covers, September 12, 2024

Type the lyrics in another language and a cover sings them in a new vocal: one of Suno's voices, or, with Suno Voices, the account owner's own verified voice. The recording the artist made stays behind.

What about Udio, Lyria, ElevenMusic and Meta?

UDIO

A licensed platform, due in 2026

After its October 29, 2025 settlement with Universal Music Group.

LYRIA 3

30-second tracks from text and images

In the Gemini app since February 18, 2026.

LYRIA 3.5

Full songs, vocal or instrumental

In Flow Music since July 29, 2026, the Gemini app and API since September 4, 2026.

ELEVENMUSIC

New compositions

On Music v2.5 since September 11, 2026; its audio reference "does not copy or remix the uploaded audio".

STABLE AUDIO 3.0

Music and sound effects

Stability AI, May 20, 2026, "trained on fully licensed data."

META

Songs in Instagram direct messages

Reported October 4, 2026; Meta's last named music model, JASCO (June 2024), makes music from text.

None takes a recorded song in to carry it into another language. Google: "Music generation with Lyria 3 is designed for original expression, not for mimicking existing artists." ElevenLabs' dubbing product is built for speech:

ElevenLabsElevenLabsElevenLabsTHEIR DOCS, READ OCTOBER 30, 2026

Editing the translation means their old model.

On ElevenLabs' current Dubbing v2, in alpha, editing the translation and regenerating the audio through the API is for their Enterprise plans only. Their full editor, Dubbing Studio, is "only available via the v1 model" and "in maintenance mode", in their words. Dubbing v1 is the model our 418-pair study measured, and its output is mono.

+27.8%

FAMILIAR LABS CLOSER TO THE REAL SPEAKER · ALL 11 LANGUAGES

261%

ELEVENLABS MORE BACKGROUND ERROR · THE MUSIC, THE AMBIENCE

+243.7%

FAMILIAR LABS ON LAUGHTER AND REACTION SHAPE

ELEVENLABS DUBBING V1 · 418 PAIRED OUTPUTS · 11 LANGUAGES · AUGUST 5, 2026Read the study Their docs

What happens to the original singer?

SUNO COVERS
WHAT GOES IN
A track and its lyrics
WHAT COMES OUT
The same melody in a new style
WHO SINGS
A new vocal
UDIO
WHAT GOES IN
A prompt
WHAT COMES OUT
A new song
WHO SINGS
A new vocal
LYRIA 3.5
WHAT GOES IN
Text or an image
WHAT COMES OUT
A new song with lyrics
WHO SINGS
A new vocal
ELEVENMUSIC
WHAT GOES IN
A prompt
WHAT COMES OUT
A new composition
WHO SINGS
A new vocal
SONG DUBBING
WHAT GOES IN
The artist's recording
WHAT COMES OUT
The same song in another language
WHO SINGS
The artist, as recorded

A generator cannot give you the voice you have known for years, the breath before the chorus, the catch on the high note.

What does Song Dubbing do?

DUBBING, NOT GENERATING

Song Dubbing carries the song an artist already made, in that artist's voice, into another language.

The lyric is translated for meaning, rhyme and syllables, the artist or their team approves every line, and the album's music videos are dubbed with it. Nothing is sung again; the original performance is carried over: the first true song dubbing (the full definition).

Which one should you use?

AN AI MUSIC GENERATOR

  • A new song, or a new style for your own.
  • Background music where nobody listens closely.

SONG DUBBING

  • Listeners understand the song they already love.
  • The singer stays the singer.

Our CEO's side: The Opposite of AI Music Generation. Song Dubbing is for record labels and their artists; there is no upload for songs. Email the team.

QUESTIONS

Can Suno translate a song into another language?

Suno Covers keeps a song's melody, remakes it in a new style, and can sing lyrics you type in another language. The voice is a new vocal; the original recording does not carry over.

Can Udio translate a song?

We found no song translation feature in Udio's public announcements. It is building a licensed service to create, stream and share music, due in 2026; what it makes is a new song.

Can Google's Lyria translate a song?

No. Lyria 3 and Lyria 3.5 make new music from text and images; they do not take a recorded song in.

Does Familiar Labs generate music?

No. Song Dubbing carries the recording an artist already made, in the artist's own voice, into another language, with the artist approving every translated line. Dubbing, not generating.

REFERENCES

  1. [1]Suno, Introducing Covers, September 12, 2024
  2. [2]Suno, Voices
  3. [3]Music Business Worldwide, Universal Music settles Udio lawsuit, October 30, 2025
  4. [4]Google, Gemini can now create music (Lyria 3), February 18, 2026
  5. [5]Android Authority, Lyria 3.5 comes to the Gemini app, September 4, 2026
  6. [6]Google AI Studio, Lyria 3.5 model page
  7. [7]Music Business Worldwide, Google upgrades Lyria (Lyria 3.5), July 30, 2026
  8. [8]ElevenLabs, Introducing Music v2.5, September 11, 2026
  9. [9]ElevenLabs, Eleven Music documentation
  10. [10]Stability AI, Meet Stable Audio 3.0, May 20, 2026
  11. [11]Social Media Today, Instagram adds AI songs to DMs, October 4, 2026
  12. [12]Meta AudioCraft, JASCO documentation (released June 2024)
  13. [13]Song Dubbing: The Same Singer, in Your Language

Multimodal Translation

Familiar Labs builds generationally better dubbing with a Multimodal Localization Foundation Model (MLFM), the first foundation model for multimodal translation: one model decides what to localize across text, audio and visual, and shows you the choices. Your voice and lips, carried into new languages. They're a team of 7 with PhDs, backed by Samsung, RedBird, Gold House. Our mission