20 free, copy-paste prompts for Google Lyria 3.5 — Google DeepMind's latest AI music model. Section-level control, precise BPM and key, full vocals and lyrics. Access free via Google Flow Music. No signup needed here.
The Google Lyria 3.5 prompt generator on this page gives you 20 professionally written, copy-ready prompts for Lyria 3.5 — Google DeepMind's most advanced AI music model, released July 29, 2026, and available for free inside Google Flow Music.
Lyria 3.5 is the first widely accessible AI music generator with true section-level control: you can label [intro], [verse], [chorus], [bridge], and [outro] in your prompt and Lyria follows that structure precisely. Add exact BPM, musical key, specific instruments, vocal style, and an artist reference — and Lyria produces a polished, full-length track that actually sounds like what you described.
Every prompt below is structured for Lyria 3.5's section-based system. They also work — with minor adjustments — in Suno V4, Udio, and ElevenLabs Music.
Lyria 3.5's section system is its biggest differentiator. Use this structure:
Click any prompt to copy — paste into Google Flow Music and generate
[intro: 8 bars, acoustic guitar] [verse: melancholic indie pop, 84 BPM, G major, acoustic guitar fingerpicking, warm female vocals] [chorus: soaring, layered harmonies, driving drum kit, emotional release] [bridge: stripped to voice and guitar] [outro: fade on guitar] — bittersweet, nostalgia, late summer, Phoebe Bridgers influence
[intro: low strings, building tension] [verse: full orchestra swells, D minor, 130 BPM, brass fanfare, thundering timpani] [chorus: heroic choir, soaring violins, maximum orchestral impact] [outro: gentle piano resolution] — no lyrics, film trailer quality, Hans Zimmer influence, emotional journey
[intro: funky bass groove, 120 BPM, F major] [verse: four-on-the-floor kick, wah guitar, falsetto male vocals, tight horn stabs] [chorus: massive disco hook, group backing vocals, glittery production] [bridge: percussion breakdown] — joyful, dancefloor energy, Daft Punk meets Bee Gees
[intro: talking drum pattern, marimba] [verse: 105 BPM, A major, thumb piano melody, punchy bass, bright female lead vocals in Yoruba-English] [chorus: call-and-response, celebratory crowd energy, layered percussion] [outro: percussion solo fade] — joyful, West African rhythm, Burna Boy production quality
[intro: distant industrial drone] [verse: 128 BPM, C minor, distorted synth bass, rigid drum machine, half-spoken male vocal] [chorus: full distorted wall of sound, urgent, cinematic tension] [bridge: silence then single piano note] — anxious, cyberpunk atmosphere, Nine Inch Nails influence, no clean resolution
[intro: solo fingerstyle guitar, 70 BPM, D major] [verse: intimate male baritone, narrative folk lyrics about leaving home, light brush percussion] [chorus: gentle harmony, banjo joins, emotional warmth] [outro: voice trails, single guitar chord] — Townes Van Zandt storytelling, campfire intimacy, no production polish
[intro: Spanish guitar riff, 96 BPM] [verse: bilingual Spanish-English lyrics, warm male lead, reggaeton dembow rhythm under melodic pop verse] [pre-chorus: tension build, strings enter] [chorus: massive melodic hook, full production, stadium feel] — confident, romantic, Bad Bunny meets J Balvin modern production
[intro: ambient piano, 60 BPM, trap hi-hat pattern] [verse: 808 bass, whispered male falsetto, auto-tune lightly applied, confessional lyrics] [chorus: emotional break, full 808s, layered harmonies] [bridge: stripped falsetto only] — late-night vulnerability, Frank Ocean adjacent, moody atmospheric reverb
[intro: distorted guitar riff, 115 BPM, E minor] [verse: driving rhythm guitar, gritty male vocal, restrained dynamics] [chorus: massive wall of guitar, powerful release, singalong hook] [bridge: feedback and breakdown] [outro: long guitar fade] — angst, raw energy, Radiohead meets Foo Fighters production, no synths
[intro: steel pan melody, 80 BPM, G major] [verse: classic reggae one-drop rhythm, skank guitar, warm male vocal with patois inflection, upright bass walk] [chorus: relaxed open melody, backing harmonies] [bridge: dub breakdown, echo effects] — sun-soaked, Bob Marley influence, live-band feel
[intro: electronic stinger, 130 BPM, F# major] [verse: punchy 808 bass, bright synth lead, girl group rap verse in Korean] [pre-chorus: tension riser] [chorus: massive melodic hook, layered harmonies, choreography energy] [bridge: slower emotional moment] — polished idol production, BLACKPINK and aespa influence
[intro: warm Rhodes piano chord, 68 BPM] [verse: E-flat major, breathy female vocal, lazy jazz bass, finger snaps, light brushed drums] [chorus: fuller arrangement, organ swells] [bridge: vocal ad-lib showcase] — intimate, sophisticated, Erykah Badu and H.E.R. influence, unhurried feel
[intro: four count stick click, 160 BPM] [verse: power chords, shouted male vocal, driving bass, minimal but punchy] [chorus: gang vocals, maximum energy] [outro: abrupt cut-off] — 90 seconds total, raw, unpolished, Rancid meets The Clash, no production gloss, live recording vibe
[intro: wall of reverbed guitar, 90 BPM, A major] [verse: hushed female vocal buried in the mix, dreamy, indistinct lyrics, slow-moving harmony] [chorus: swelling guitar wash, barely distinguishable melody] [outro: infinite reverb decay] — My Bloody Valentine influence, texture over structure, emotional but vague, no sharp dynamics
[intro: tight bass line, 102 BPM, D minor] [verse: slap bass, wah guitar, punchy horns, call-and-response male vocal] [chorus: full horn section blast, danceable, infectious] [bridge: guitar solo over groove] — James Brown and Bruno Mars influence, live-band energy, nothing electronic
[full track: 50 BPM, sustained drone pads, singing bowls, sparse slow piano notes] — no structure, no build, no lyrics, binaural texture, field recording of rain underneath, Brian Eno influence, 4-minute continuous texture, floating, meditative, no percussion
[intro: banjo run, 88 BPM, B-flat major] [verse: acoustic flat-pick guitar, upright bass thump, warm male tenor, fiddle countermelody] [chorus: group harmony, full band, barn-dance energy] [bridge: fiddle breakdown solo] — Appalachian tradition, Doc Watson influence, no electric instruments, live porch sound
[intro: 16-bar hi-hat build, 132 BPM] [verse: deep synth bass, filtered house chords, processed vocal chop] [pre-chorus: tension riser, white noise sweep] [chorus: full drop, four-on-the-floor kick, euphoric synth lead] [outro: gradual strip to kick only] — underground club energy, no lyrics, Chicago and Berlin house influence
[intro: steel guitar lick, 100 BPM, F major] [verse: breathy Japanese female vocal, warm 80s production, Fender Rhodes, walking bass] [chorus: lush chord progression, background 'city at night' feeling] [bridge: saxophone solo] — Mariya Takeuchi influence, nostalgic, polished 80s Tokyo aesthetic, bittersweet
[intro: Hammond organ chord, 96 BPM, E-flat major] [verse: powerful female lead soloist, testifying, emotional] [chorus: full SATB gospel choir, hand claps, call-and-response] [bridge: choir alone, a cappella moment] [outro: choir and organ build to peak] — Southern Baptist tradition, Kirk Franklin influence, uplifting, goosebump moment
| Feature | Lyria 3.5 | Suno V4 | Udio | ElevenLabs Music |
|---|---|---|---|---|
| Section control | ★★★★★ [verse]/[chorus] labels | ★★☆☆☆ Auto-generates | ★★★☆☆ Partial | ★★★☆☆ Moderate |
| Key/BPM accuracy | ★★★★★ Very precise | ★★★★☆ Good | ★★★☆☆ Variable | ★★★★☆ Good |
| Vocal quality | ★★★★☆ Natural, articulate | ★★★★☆ Consistent | ★★★☆☆ Experimental | ★★★★★ Highest clarity |
| Max length | 3 minutes | 4 minutes | ~3 minutes | ~4 minutes |
| Free tier | Yes — Google Flow Music | Yes — 50 credits/day | Yes — 40 credits/day | Yes — limited |
| Best for | Structured songs needing precise section control and key accuracy | Genre-accurate full songs, reliable results | Experimental, genre-fusion outputs | Vocal-forward tracks, voice quality priority |
Google Lyria 3.5 is Google DeepMind's latest AI music generation model, released July 29, 2026. It generates full songs with vocals, instruments, and lyrics from a text prompt — accessible through Google Flow Music (free tier with daily usage limits, extended access via Google AI subscription). Lyria 3.5 introduced structured section-level generation (you can label intro, verse, chorus, bridge, and outro separately), tighter lyric adherence, and precise tempo/key control — the biggest quality jump in the Lyria lineage.
Lyria 3.5 is available inside Google Flow Music — the AI music creation tool in the Google AI suite. The free tier gives you a daily credit allowance (roughly 5–8 songs per day). Google AI subscribers (part of the Google One AI Premium plan) get expanded credits and higher-quality export options. Go to flow.google.com, create a project, type your prompt, and Lyria 3.5 generates the full track in about 45 seconds.
Lyria 3.5's defining advantage is section-level control — you can explicitly label [intro], [verse], [chorus], [bridge], and [outro] in your prompt, and Lyria follows that structure with high fidelity. Suno V4 auto-generates structure (less control). Udio tends toward more experimental, unpredictable results. Lyria 3.5 also has the strongest key and tempo adherence — specifying '84 BPM, G major' reliably produces a track in that exact key and tempo. For structured pop, folk, or classical compositions, Lyria 3.5 is currently the most controllable AI music model.
Lyria 3.5 responds best to prompts with: (1) Section labels — [intro], [verse], [chorus], [bridge], [outro] with descriptions for each; (2) Exact BPM and key — '84 BPM, G major' gives Lyria a precise musical target; (3) Specific instruments — 'nylon string guitar', 'Hammond organ', 'talking drum' rather than generic terms; (4) Vocal style and language — specify gender, register, and lyric language; (5) Artist or era reference at the end as a style anchor. The 20 prompts above use this structure throughout.
This depends on your Google AI subscription tier. Google's current policy allows personal, non-commercial use on the free tier. Google AI Premium subscribers receive commercial rights for tracks generated within their subscription. Always check the current Google Flow Music terms at flow.google.com/terms — music AI licensing is a rapidly evolving area and terms have changed across the Lyria family since v1 in 2023.
Yes, with adjustments. The section labels [intro], [verse], [chorus] are native to Lyria 3.5 — remove them for Suno V4 and Udio, which don't parse section structure from prompts. The genre, mood, BPM, instrument, and artist-reference components transfer directly to Suno and Udio. For ElevenLabs Music, focus on the vocal style and genre description — it prioritises voice quality over structural arrangement.