Google Flow Music review (2026): honest scoring on Lyria 3.5 vocals, lyrics, controls, pricing, watermarking, and who the AI music studio actually fits.
Google Flow Music is a polished, capable AI music studio, and Lyria 3.5 is a real step up — noticeably more natural vocals and lyrics, richer melodies, and welcome control over tempo and duration. It is strong at its one job, easy to access through a Google account, and free to upgrade. The honest limits are scope and provenance: beyond a song, a music video, and cover art it does no captioned social short-form, multi-format posts, or cross-platform publishing, and every track is SynthID-watermarked into a shifting AI-music disclosure landscape. Judge it as a music tool, not a content engine.
Google Flow Music is Google's dedicated AI music studio — a browser app at flowmusic.app and an iOS app — powered by DeepMind's Lyria model. On July 29, 2026 Google made Lyria 3.5 the default engine, pitching it as a broad quality upgrade: more natural and complex melodies, higher-quality lyrics with better prompt adherence, more realistic and emotionally nuanced vocals with cleaner pronunciation, and direct control over tempo and duration. For existing users it arrived free, with no new tier and no waitlist.
That cadence matters. Lyria 3 shipped in February 2026, Lyria 3 Pro followed in late March with tracks up to roughly three minutes, and Lyria 3.5 now refines the core output — three milestones in a year signal Google treats AI music as a standing product line. The same model also powers music in Gemini, YouTube Shorts' Dream Track, and Google Vids, so Flow Music is the front door to a broader Lyria stack.
This review is for anyone deciding whether Flow Music earns a spot in their workflow. I run a competing content product, Kompozy — but Kompozy does not generate music, so this is not a head-to-head, and I am not going to invent weaknesses to sell you something. Flow Music is good at a real, narrow job. The honest work here is mapping where it shines, where the watermarking and music-only scope show, and where it simply stops — because a music studio and a content engine are not the same tool.
Google Flow Music generates original songs. You give it a text prompt, an image, or your own lyrics, and Lyria 3.5 composes a full track — vocals, lyrics, and instrumentation — across a wide range of genres and global languages, up to roughly three minutes long. The Lyria 3.5 update specifically improved vocals, lyric quality, melodic structure, and creative control, letting you steer tempo and duration more directly. It runs tied to a Google account, and every track is embedded with SynthID, Google's imperceptible watermark, so the audio can be identified as AI-generated or AI-edited. Access is Google-ecosystem-shaped: a free plan with daily credit top-ups and limited concurrent generations, paid Flow Music tiers (reported around $8, $24, and $64 per month for more songs), and equivalent access bundled into Google AI Plus, Pro, and Ultra subscriptions. It is a music generator — in Spaces it can spin up a music video and cover art for a track — but it is not a social-video editor, a post-graphics tool, or a publisher; it hands you the song and its assets.
The clearest fit is a creator or producer who needs to generate original music fast: a short-form creator scoring Reels and TikToks with an AI hook, a hobbyist songwriter experimenting with genres and lyrics, or someone already paying for Google AI who gets the studio bundled in. It suits people who value the newest AI vocal quality and an easy, no-setup on-ramp through a Google account. It is not for someone whose real need is generating captions, carousels, threads, or captioned social video, or publishing across platforms — Flow Music does none of that, and a creator with that bottleneck will get a great song and still be stuck on everything after it.
| Dimension | Score | Why |
|---|---|---|
| Vocal quality (Lyria 3.5) | 4.3 / 5 | The headline upgrade — vocals are noticeably more realistic and expressive with cleaner pronunciation, though results still vary by prompt. |
| Lyrics and prompt adherence | 4.0 / 5 | Improved prompt adherence and structural awareness make for more coherent, on-topic lyrics than earlier versions. |
| Musicality and melodic structure | 4.1 / 5 | Richer, more natural melodies that hold up across a full track, addressing the old mechanical-loop criticism. |
| Creative control (tempo, duration, steering) | 3.9 / 5 | Direct tempo and duration control is a genuine addition; fine-grained editing is still limited compared with a DAW. |
| Genre and language range | 4.2 / 5 | Broad genre coverage and global-language support give it real range from a single prompt. |
| Ease of use and access | 4.4 / 5 | Web and iOS, a free plan, and a Google-account sign-in make it about as frictionless as AI music gets. |
| Pricing and value | 4.1 / 5 | A free tier to try, affordable paid plans, and inclusion in Google AI subscriptions make it good value for the music job. |
| Rights, watermarking, and provenance | 3.7 / 5 | SynthID on every track is good provenance, but AI-music disclosure and monetization rules are unsettled, so commercial use needs care. |
| Scope beyond music (production + publishing) | 2.3 / 5 | Beyond a song, its music video, and cover art, no captioned social short-form, multi-format post fan-out, brand-voice governance, or scheduling. It stops at the track. |
Flow Music's pricing is straightforward and, for a music tool, fair. A free plan with daily credit top-ups and a couple of concurrent generations lets you try Lyria 3.5 before paying — and, notably, Lyria 3.5 itself was a free upgrade for existing users, with no new tier or waitlist. Paid Flow Music plans are reported around $8/month (Starter), $24/month (Plus), and $64/month (Member), scaling how many songs you can make. The neat part of Google's model is bundling: Google AI Plus, Pro, and Ultra subscribers get the equivalent Flow Music tier included, so if you already pay Google for AI, the music studio is largely free at the margin.
Measured against standalone AI-music rivals, that pricing is competitive, and the free-upgrade path means you are not paying extra for the latest model. The caveat worth flagging is provenance rather than cost: every track is SynthID-watermarked, and platforms are increasingly badging or de-monetizing AI music, so the "value" of a track depends on how you intend to use it. For monetized or brand work, factor in disclosure and keep a licensed option in reserve.
The honest framing on value is the same as the scope point. Flow Music prices the music step well, but the music step is one part of a content workflow. Your total cost to actually produce and publish video is Flow Music plus whatever edits, captions, and schedules it — so weigh the subscription as the soundtrack line item, not the whole pipeline. Always confirm current pricing on Google's live page, since tiers move.
| Use case | Fit | Why |
|---|---|---|
| Generating an original song from a prompt or lyrics | Strong | This is exactly what Flow Music and Lyria 3.5 are built for, and the vocal upgrade makes results more usable. |
| Making AI hooks, intros, or jingles for short-form video | Strong | Tempo and duration control plus quick generation suit scoring a track to a specific slot. |
| Experimenting across genres and languages | Strong | Broad genre and global-language coverage make it easy to explore styles from one prompt. |
| Getting AI music bundled with a Google AI subscription | OK | If you already pay for Google AI, the equivalent Flow Music tier is included, which is good value — but access is ecosystem-tied. |
| Producing the social video itself (not just the music) | Weak | Flow Music can make a music video for a track, but not the captioned, per-platform social video creators post — no clip formats or reframing. |
| Generating captions, carousels, or posts for a clip | Weak | Flow Music generates no captions, carousels, or post copy; that is a separate job. |
| Publishing finished videos across social platforms | Weak | There is no scheduler or publishing layer — posting is manual in other tools. |
| High-stakes monetized use with clear rights | OK | SynthID provenance helps, but AI-music disclosure and monetization rules are unsettled, so confirm each platform's policy. |
Kompozy belongs in this list with an asterisk, because it is not competing with Flow Music for the same click — and it does not generate music. Flow Music is where a song gets made: Lyria 3.5 turns a prompt, image, or lyrics into a full track with vocals. Kompozy is the next stage — it takes a finished track and turns it into published content, building the video around it, generating captions, quote cards, carousels, and Persona posts in your brand voice, reframing per platform, and scheduling across TikTok, Reels, Shorts, LinkedIn, X, and the rest of nine destinations.
So the honest positioning is a handoff, not a head-to-head. If your whole need is "make this song," Flow Music is the right tool and Kompozy has nothing to add to the audio. The moment your need becomes "make this song and then turn it into a week of posts everywhere," Flow Music stops and Kompozy starts. A clean way to run both: generate the track in Flow Music, then drop the export into Kompozy to fan it into a Clipped Short, a quote card, a carousel, and supporting text — produced, captioned, and scheduled in one pass. And because Kompozy is source-agnostic, you can swap a SynthID-watermarked track for a licensed one on a monetized post without rebuilding the video.
For generating original AI music, yes. Lyria 3.5 brings more natural vocals and lyrics, better melodies, and tempo and duration control, the studio is easy to access, and it is free to upgrade. Judge it as a music tool, not a content engine — it does no editing or publishing, and every track carries a SynthID watermark you should plan disclosure around.
Lyria 3.5 is Google DeepMind's AI music model, made the default in Google Flow Music on July 29, 2026. It improves musicality, lyrics, vocals, and creative control over earlier versions, and arrived as a free upgrade for existing users with no new tier or waitlist.
All three generate songs from prompts, but Suno and Udio are standalone tools currently in label litigation over training data, while Flow Music is Google's studio powered by Lyria 3.5, bundled into Google AI subscriptions, and watermarked with SynthID. Vocal and lyric quality vary by tool and prompt — try each on your own material.
Flow Music has a free plan with daily credits, plus paid tiers reported around $8, $24, and $64 per month for more songs; Google AI Plus, Pro, and Ultra include the equivalent tiers. Lyria 3.5 is a free upgrade. Confirm current numbers on Google's live pricing page, since plans change.
Every track is SynthID-watermarked as AI-generated, and platforms increasingly badge or de-monetize AI music. Check each platform's policy and Google's terms for your specific use, plan disclosure, and for high-stakes commercial posts keep the option to swap in a licensed track.
It can generate a music video for a song inside Spaces, but it is not a general video editor and it does not add social captions, build other post formats, or publish anywhere. For captioned per-platform short-form, supporting posts, and cross-platform publishing you pair it with a content engine like Kompozy.
Yes, and it is the natural setup. Generate your track in Flow Music, then bring the export into Kompozy to build the video around it, caption it, reframe it per platform, fan it into supporting posts in your voice, and publish across nine platforms.
Anyone whose bottleneck is producing or publishing content rather than making a song. Flow Music will hand you a great track and stop; it has no path to captions, multi-format posts, or cross-platform scheduling, and it does not produce the captioned, per-platform social video or supporting posts you publish.
See Google Flow Music vs Kompozy comparison → · Get Started →