Verses, choruses and a bridge in one generation, sung on your lyrics or kept instrumental.
Each track below is a single Lyria 3 Pro generation in Renoise; the cover art was made separately in Renoise. Hover over a track or tap its speaker to listen.
One generation returns a whole track, not a 30-second loop. There is no length setting: write “about 90 seconds” or “a full 3-minute song” in the prompt. Length is approximate; in a Renoise test, a 40-second request came back at 63 seconds.
Lyria 3 Pro follows song structure. Ask for an intro, two verses, a chorus that comes back and a bridge before the last chorus; Google says the model builds the arrangement around those sections instead of looping one idea.
Paste your lyrics into the prompt under tags like [Verse] and [Chorus], then describe the voice: a soft female vocal, a gritty male lead, a choir on the hook. Give only a theme and, per Google, Lyria writes the words for you.
Write “instrumental, no vocals” and Lyria 3 Pro returns a clean bed for a product ad, a vlog or a trailer: lo-fi beats, cinematic strings, upbeat synth-pop. Set the mood and tempo, then lay it under the clip you made in Renoise.
Add one guide image, such as album art, a photo or a still from your video, and Lyria 3 Pro carries its mood into the music. It is the only input besides your text: Lyria takes no reference audio or video.
Lyria 3 Pro is Google DeepMind’s music model, released on March 25, 2026. It writes complete songs up to about 3 minutes, with verses, choruses and a bridge, sung on your lyrics or instrumental.
Renoise has run it since October 13, 2026, on one subscription with 30+ other models, in Audio, on the Canvas, or from Claude Code, Codex and OpenClaw.
Need exact lyrics sung by a chosen lead? Try Mureka V9. Making the video? See the AI music video guide. Want the Agent to write the song and pick the model? Try the AI song generator.
Specs as available in Renoise, checked November 6, 2026.
Open Audio in Renoise, or add an audio node on the Canvas, and choose Lyria 3 Pro from the model menu.
Name genre, mood, tempo, instruments and the length you want. Put lyrics under [Verse] and [Chorus], or write “instrumental”.
Optionally add one guide image and generate. At 4 credits a song, try two or three takes.
From Google’s first 30-second Lyria 3 tracks to full songs in Renoise.
Google brings Lyria 3 to the Gemini app as a beta that makes 30-second tracks from a prompt or a photo.
Google announces Lyria 3 Pro with tracks up to 3 minutes, in Vertex AI (preview), AI Studio and the Gemini app.
Google ships a newer Lyria 3.5 in its Flow Music app. Renoise does not offer it yet.
Lyria 3 Pro replaces the 30-second Lyria 3 Clip in Renoise, in Audio, on the Canvas, the API, the CLI and MCP.
Sources: Google blog, Google Workspace Updates, Gemini API docs, Renoise model list. Last updated
Renoise ran the 30-second Lyria 3 Clip until October 13, 2026. Here is what moving to Lyria 3 Pro changes.
Lyria 3 Clip always returned about 30 seconds. Lyria 3 Pro goes up to about 3 minutes, so one generation covers a whole song or a full video edit.
Google built Lyria 3 Pro to follow sections. Name the intro, verses, choruses and bridge, and the song develops instead of repeating one loop.
A Lyria 3 Pro song is 4 credits; a Lyria 3 Clip was 2. Short beds still cost 4, so ask for the length you need.
New music tasks named Lyria now run on Lyria 3 Pro. Your earlier Clip results stay in your library.
Three music models in Renoise Audio, compared on what Renoise exposes for each, not on each vendor’s full spec sheet.
| In Renoise | Mureka V9 | ElevenLabs Music | |
|---|---|---|---|
| Song length | Up to ~3 min, set in the prompt | ~2.5–3.5 min, no length setting | ~3–3.5 min, no length setting |
| Vocals and lyrics | Optional; lyrics in the prompt | Always sung; lyrics required | From the prompt; no lyrics field |
| Instrumental only | Yes, ask for it | No | Not guaranteed |
| Other input | 1 guide image | None; pick a male or female lead | None |
| Credits · 1 song | 4 | 9 | 90 |
| Best for | Songs and scores from a brief | Singing exact lyrics | Text-only, full-length tracks |
Renoise model list and credit estimates (checked 2026-11-06).
As of October 2026, 100,000+ creators in 170+ countries and regions use Renoise AI, an AI video, image and audio platform, to generate 300,000+ images and videos with 30+ models.
Source: Renoise internal data, March–October 2026. Figures are rounded down and updated quarterly. Last updated
Real feedback from creators who use Renoise AI to make videos, images and ads.
Renoise is pretty easy to use, with a clean and smooth interface. The results from Seedance 2.5 also feel noticeably smoother in action sequences.
Renoise Ai is a good tool for making ai videos and photos. it puts several Ai models in one place, so you don't need to use many different apps.
あまりAIツールを使ったりしたことなかったですが、キャラクターの固定や動画を続いて作成できて使いやすいです。これからどんどん進んでいくアップデートが楽しみです。
Renoise has been a great platform for AI creativity. The tools are easy to use, the experience is smooth, and it makes creating content more enjoyable and efficient. I really appreciate the work the team is doing and look forward to seeing more improvements. Highly recommended!
What I like about Renoise AI compared to other platforms: Fast generation speed: Slow generation takes up way too much time, but Renoise AI is surprisingly about 30% faster. Intuitive and user-friendly UI: Cheaply made platforms have terrible UIs that make me not even want to log in, but Renoise is very clean and convenient. Remembers my previous settings: It helpfully remembers my last workspace settings. Some odd platforms force you to re-adjust everything every single time you generate a video, which is extremely frustrating. Overall preference: For all these reasons, Renoise AI is definitely my preferred platform. Thank you!
Renoise is your go to one stop all for ai creators. all view and use all models in one place, for any type of content you have your mind on!
I have a creat experience with the app. Consistent characters but also unique. THere were some issues with storyline continuity and non-typical humanoid characters.
One Renoise plan unlocks Lyria 3 Pro, the other music models and every video and image model, billed per generation from your plan credits.
For first-time AI content creators
Billed $168 annually
Upgrade PlanSave $72/year
For creators shipping content every week.
Billed $504 annually
Upgrade PlanSave $216/year
For studios and pros producing at commercial scale.
Billed $1680 annually
Upgrade PlanSave $720/year
For first-time AI content creators
Billed monthly
Upgrade PlanFor creators shipping content every week.
Billed monthly
Upgrade PlanFor studios and pros producing at commercial scale.
Billed monthly
Upgrade PlanRenews automatically each year until canceled.
Renews automatically each month until canceled.
Yes. Lyria 3 Pro has been live in Renoise since October 13, 2026, in Audio, on the Canvas, and through the Renoise API, CLI and MCP. It replaced the 30-second Lyria 3 Clip, which no longer takes new tasks. Google announced Lyria 3 Pro on March 25, 2026.
Up to about 3 minutes. There is no duration setting in Renoise: describe the length in the prompt, such as “a 60-second instrumental” or “a full 3-minute song”, and treat it as a target, not a promise. For an exact cut, ask for a little more and trim it in your edit.
Yes. Write the lyrics in the prompt, split into sections like [Verse], [Chorus] and [Bridge], and describe the voice you want. Keep the lyrics separate from the style notes. To have every line sung by a chosen male or female lead, use Mureka V9 in Renoise, which takes the lyrics as a separate field; it may add ad-libs of its own.
Yes. Say “instrumental, no vocals” in the prompt and Lyria 3 Pro returns music without singing, which suits background music for ads, vlogs, games and trailers. Add the mood, tempo and instruments you want, and the length of the video you are scoring.
It accepts one optional guide image, such as album art or a frame from your video, to set the mood. It does not take reference audio or video, so it cannot continue, remix or match an existing track. Describe the sound you want in words instead.
In Renoise, a Lyria 3 Pro song costs 4 credits, the same whatever length you ask for. For comparison, Mureka V9 is 9 credits and ElevenLabs Music is 90 credits a song. Credits come from your Renoise plan; see pricing for plans.
Not in Renoise. Lyria 3 Pro is billed per song from the credits in your Renoise subscription, at 4 credits a song, like every other model. Google’s own apps have their own plans and limits, which Renoise does not set.
Yes. You own the output, and commercial use follows your Renoise plan, under our Terms of Service. Do not prompt for a named artist’s voice or a copyrighted song. Google embeds a SynthID watermark in every Lyria 3 Pro track.
No. Do not ask for a named artist’s voice or style; Google blocks prompts that request specific artist voices, and Renoise’s terms forbid using someone’s voice without permission. Describe the sound instead: era, genre, instruments, tempo and the character of the vocal.
Yes, through Renoise. Call `lyria-pro` (aliases `lyria` and `lyria-3-pro`) with the same Renoise API key, CLI and MCP you use for video and image models, and send a text prompt plus at most one image. Google also offers Lyria 3 Pro to developers in Vertex AI and AI Studio.
Not yet. Google released Lyria 3.5 in its Flow Music app on July 29, 2026, with improved vocals and variable song lengths, and later added it to the Gemini API. Renoise runs Lyria 3 Pro today, the model this page describes.
Lyria is Google DeepMind’s family of music models: Lyria 3, Lyria 3 Pro, Lyria 3.5 and Lyria RealTime. Renoise does not train music models; it integrates Lyria 3 Pro next to Mureka V9, ElevenLabs Music and Seed Audio 1.0, plus 30+ video and image models.
Write a song or score a video with Google’s Lyria 3 Pro, next to every video model.