Complete song prompts — style, arrangement, lyrics

Getting Started

Lyricoro generates complete, copy-paste-ready prompts for Suno. Each generation produces two text blocks:

How Generation Works

Pick a genre family — it controls everything: instruments, vocal style, BPM, arrangement structure. Lyrics are filled from a built-in library or generated live by AI. The result is two copy-paste text blocks formatted for your chosen platform (Suno, Udio, ACE-Step, or Generic).

Step by Step

Options Explained

Concept

Describe the theme, mood, or scene for your song — a few words or a short phrase. Examples: "heartbreak", "summer party", "ночная тоска", "school night out".

When you provide a concept, the engine analyzes its emotional tone and selects lyrics that match that mood. A sad concept pulls melancholic lines, a party concept pulls upbeat ones — all sections stay tonally consistent. Works cross-language: an English concept like "summer vacation" will correctly filter Russian lyrics, and a Russian concept like "ночная тоска" will filter English lyrics.

Family

A genre template that controls everything about the sound: vocal pool, bass and lead instruments, BPM range, arrangement vocabulary, and production vibe. "mixed" is a family in its own right, drawing on a wide genre list — it does not roll a die and pick one of the other 25. Leave it selected and type a concept, and the concept decides which family fits.

Lyrics Mode

• Tags Only — section tags with production direction only. No actual lyrics. Use this when you want to write lyrics yourself but need the arrangement structure and style.
• Lyrics — fills each section with lyrics from a built-in library of 18,900+ curated entries. Instant, free, and genre-matched. A recency system prevents repeats across consecutive generations.

Language

Controls the lyrics language. EN for English, RU for Russian (modern, colloquial, not literary). Auto uses the family's default — Russian families default to Russian, everything else to English.

Vocal

Selects the vocal style tag in the Style Box. Male picks male vocal descriptions, Female picks female, Duet picks duo/group styles. Auto picks randomly from whatever the family offers.

Ad-libs

Inserts vocal cues, beat drops, and production tags into the lyrics. These are placed by meaning — not randomly. A whispered tag before a quiet verse, a beat drop before the chorus, a stutter on the hook word. Genre-aware: hookah-rap gets different ad-libs than drill or disco.

Works in Lyrics mode, where placement is rule-based from the section and its mood. No effect in Tags Only mode, which has no lyrics to annotate.

Duet, M+M, F+F

Picking a duet does two things. The style box asks for two singers, and the lyrics are marked up so Suno knows who sings what: the verses trade between the two voices, the pre-chorus stays with the verse it follows, and every chorus is marked for both. Whole sections alternate rather than individual lines — two singers delivering different words at the same moment tends to collapse into mush, so the singers meet on the hook instead.

Sung: Full song / Chorus only / No vocals

One control with three settings, because they are three answers to the same question and only one can be true at a time. It used to be two separate checkboxes that switched each other off.

No vocals

A track with no singing at all. No lyrics are written, every section becomes an [Instrumental ...] section, and every vocal description is stripped out of the style box — including genre names like "Vocal Jazz" and instrument lines like "brass section answering vocal phrases". Removing those cues matters more than adding "no vocals" as a negative, because Suno reads the style box left to right and an early positive cue overrides a later negative.

Sections are renamed to match: verses become Themes, the chorus becomes the Main Hook, the bridge becomes a Breakdown. Each one keeps its production note, which with no words is the only direction it carries. There is no quality score, because nothing was written to score.

This is not the same as Solo Break below.

Chorus only

The chorus is the only thing sung. Verses, pre-choruses and the bridge all become [Instrumental ...] sections, so the track plays as music until the hook arrives and drops back to music after it. Common in house, post-rock and trailer music, where a single repeated vocal phrase does all the lifting.

Unlike No vocals, the style box keeps its vocal description — there is still a singer, and stripping the cue would have Suno hum the hook instead of singing it.

The quality score still describes the full set of lyrics the generator drew, not just the four printed lines. Scoring the printed lines alone would report a song with no verses and no arc as structurally broken, which says nothing about the chorus you asked for.

Mixed RU/EN

A bilingual song: the verses stay in the language you picked and the chorus is sung in the other one. This is how a lot of Russian pop and dance music is actually written — the verse carries the story in Russian, the hook lands in English.

The hard part is making the two halves belong to the same song. Lyricoro draws the hook first, reads its mood, and then picks the verses from the other language's pool to match. Measured, that raises the emotional common ground between verse and chorus by two to three times against splicing two independent songs together. It is not a translation — the two halves are different words about the same feeling, which is what the real thing does too.

The style box tells Suno which half is which, so it sings each language in the right place instead of guessing.

The quality score switches to bilingual mode: each language is judged against its own vocabulary, and Theme is left out entirely. Theme measures words shared between verse and chorus, and a bilingual song shares none on purpose — reporting it would mark the format itself as a fault.

Hook First and the teaser

Hook First replaces the intro with a stripped-down teaser chorus, so the hook is the first thing anyone hears. The dropdown beside it decides what the teaser sings. Full chorus previews all of it. Line ×2, ×3 and ×4 repeat only the chorus's first line, chant-style, so one phrase is drilled in before the song starts — and the full chorus still lands whole in its own place a section later.

Solo Break

Inserts a solo/improv break after the bridge — does NOT remove vocals from the song. The break description uses the actual instruments in your style (e.g. if the style picked slap bass and Rhodes piano, the solo will reference those). Good for longer songs that need a breather between sections.

Lyrics Quality Score

Every generation gets an automatic quality rating from 0 to 10. Click the score badge below the lyrics to see the breakdown.

Cover Mode

Cover mode restyles your lyrics in a different genre — it keeps your words but wraps them in a new arrangement and generates a matching Style Box.

Paste your lyrics into the Cover tab. Lyricoro keeps your words but wraps them in a new genre's arrangement tags and generates a matching Style Box. Use [Verse 1], [Chorus] tags in your lyrics for best results. Without tags, sections are split by blank lines.

Tips for Better Results

• Edit the output — the text boxes are editable. Fix awkward lines, swap words, rewrite anything that doesn't feel right before copying to Suno.
• Try different seeds — same settings with a different seed number produces completely different output. Find a seed you like and save it.
• Use favorites/blacklist — if you always want slap bass, add it as a favorite. If you hate guttural vocals, blacklist them.
• Tags first, then fill — generate in Tags Only mode to see the arrangement, then switch to Lyrics if the structure looks good.
• The chorus is the hook — in Lyricoro, all chorus instances use the same lyrics. This is intentional. Repetition is what makes hooks stick.

I built Lyricoro because I wanted better Suno prompts than I had the patience to write by hand every time. A vague prompt gets you vague music. The difference between a forgettable generation and one that sounds like it was produced on purpose is what you tell the engine — which instruments, how the arrangement moves, what the vocals do in each section. Writing that out properly for every song is tedious work. So this does it.

It's here in case it's useful to you too. No account, no signup, nothing to install.

What You Get

Every generation produces two blocks to copy:

• Style Box — genre fusion, instruments, key, BPM, vocal direction. Goes in Suno's Style of Music field.
• Exclude styles — the negatives that keep the slop out, with the "no" stripped since the field already means exclude. Goes in Suno's Exclude styles field under More Options. That field is web only; on Suno mobile press Move into Style, which folds them back into the Style Box with the "no" restored. Kept out of the Style Box on purpose: that box reads left-to-right and truncates from the end, so trailing negatives were both the weakest thing kept and the first thing lost.
• Lyrics Box — the song section by section, each tagged with production direction: what the drums do, where the vocal drops out, how the final chorus differs from the first. Goes in the Lyrics field.

Both are editable before you copy, and in Lyrics mode any single section can be rerolled on its own without touching the rest. Output is formatted for Suno, Udio, ACE-Step, or generic.

Free, and Yours to Use

No account, no paywall, no credits. Everything Lyricoro generates for you — style blocks, arrangement tags, lyrics, and titles — is yours to use however you like, including commercial releases. Lyricoro claims no ownership of it and takes no cut.

One honest caveat: that right is non-exclusive. In Lyrics mode the words come from a shared library, so the same lines turn up in other people's songs too. Generating repeatedly inside one genre family, choruses start repeating somewhere in the first ten to thirty songs. Nobody holds exclusive rights to library text — not you, not the next person, not me.

So if you're releasing commercially, edit the lyrics. Rewrite lines, reroll sections, change the details until it reads like you. That editing is your own work, and it's what makes a song genuinely distinct rather than one draw from a shared deck. It matters practically too: distributors and streaming platforms flag duplicate content, and two releases carrying identical lyrics can end with one of them pulled.

A shout-out is welcome, though. If the lyrics did real work on a track, a mention is genuinely appreciated — "lyrics via Lyricoro" in the description, or a link to lyricoro.com wherever you list credits. It's a request and not a condition: nothing about your rights to the output changes if you skip it, and nobody is checking. It just helps other people find the tool.

How It Actually Works

Each genre family is a data set rather than a text template, which is what keeps a generation internally consistent:

• The BPM range shown for a family is the range it actually generates — read from that family's own tempo pool, not typed in by hand.
• Choose Female and the Style Box vocal is filtered to female voices and the lyrics are written from that perspective. Choose M+M and you get a same-gender duo, not a mixed one.
• Key and chord progression agree with each other — a minor key gets a minor progression.
• Every family carries its own arrangement vocabulary, so a French Touch intro doesn't read like a Phonk Drift intro.
• Leave the seed empty and each generation avoids what it just showed you. Type a seed and that seed plus the same settings reproduce the same song exactly — share links carry whatever is in the box, so set a seed first if you want the link to recreate the song rather than the setup.

Every generation is scored across eight dimensions — hooks, imagery, flow, originality, singability, language, theme, and structure. In Lyrics mode the generator builds several candidate combinations and shows you the highest-scoring one.

Why These Genres

The 27 genre families cover styles that are underserved by existing prompt tools — not just mainstream pop and hip-hop, but the genres where prompt detail matters most:

• Russian Sad Dance / Hookah-Rap — a massive scene with almost no English-language tooling
• French Nu-Disco — slap bass, Rhodes piano, French Touch production
• Trap / Drill — 808-driven, with distinct arrangement patterns for each subgenre
• Lo-Fi Bedroom / Ambient — texture-heavy, requires specific instrument and mood vocabulary
• Darkwave Synthpop — cold synths, baritone vocals, analog production aesthetic
• Latin / Guaracha — dembow rhythms, reggaeton structure, bilingual hooks
• Jazz / Swing — walking bass, brushed drums, ii-V-I harmony, big band horns, bossa nova
• Jamaican / Reggae — one drop, ska, dub, rocksteady, dancehall, lovers rock
• And more — EDM, Phonk, K-Pop, City Pop, Indie Folk, Pop Punk, Tribal, R&B, and others

They're opinionated starting points, not generic templates — each with its own vocal pool, instruments, arrangement vocabulary, and production vibe.

The Lyrics Library

18,900+ curated entries — 16,084 across 9 shared pools, plus 2,711 written specifically for individual genre families so two families sharing a pool stop writing the same song — in English and Russian. Every verse is built around concrete imagery: real places, real objects, real situations. The Russian pool references маршрутки, подъезды, and ночные фонари. The English pools reference bodegas, fire escapes, parking lots, and cold coffee. Many entries have been reworked for rhyme scheme (ABAB, AABB, ABCB, ABBA) while keeping the phrasing natural and singable. A recency system means you rarely see the same lines twice in a row.

A quality filter removes generic phrasing. If it sounds like something every other generator would produce, it doesn't make it in. The goal is lyrics with personality and specificity.

Built for Replay

Every arrangement follows a tested structure: the chorus repeats with identical lyrics so the hook sticks, sections build and release energy, and small production surprises keep it from going flat. The aim is a song that lands on the first listen.

It's still being worked on, and the output is always editable. If a line doesn't sit right, change it — that's part of the workflow, not a failure of it.

Start Generating

The Name

Lyric (Greek lyrikos) — poetry fused with music. Oro (Latin aurum) — gold. Lyricoro, golden lyrics: the idea that song words should be worth keeping, not disposable filler.

The Fine Print

Content & copyright. The built-in lyrics library was AI-generated specifically for this tool; Lyricoro does not intentionally reproduce copyrighted lyrics from published songs. Your licence to generated lyrics is non-exclusive, because the library behind them is shared — the same lines can be generated for someone else, and nobody holds exclusive rights to library text.

The library itself. Use whatever the tool generates for you, without limit. The pools behind it are a different thing — a curated collection assembled and edited over time, and not something to scrape, bulk-export, redistribute, or resell.

Text you paste in. Cover mode processes text you provide — Lyricoro does not store, cache, or redistribute user-submitted lyrics, and you are responsible for ensuring your input does not infringe third-party copyrights.

Disclaimer. All content produced by Lyricoro — lyrics, style prompts, arrangement tags, and song titles — is automatically generated and provided strictly on an "as-is" basis. The developers are not responsible for the meaning, accuracy, appropriateness, or legality of any generated text, which may contain language, themes, or imagery that is offensive, nonsensical, or culturally insensitive; none of it reflects the views or intentions of the developers. By using Lyricoro you acknowledge that you are solely responsible for reviewing, editing, and using generated content, and you assume all risk associated with publishing, performing, or distributing any output. Lyricoro is not liable for any claims, damages, or disputes arising from user-generated or tool-generated content.

Privacy. Your IP address is used for rate limiting — held in memory only as long as that check needs, and never written to disk. Usage events (which genre, which language, which platform) are stored against a salted hash of your address rather than the address itself, so visitors can be counted but not identified. All of it sits in ephemeral storage and disappears when the server restarts. The site collects anonymous page-view analytics. Messages from the contact form are emailed to me. No advertising, no tracking cookies.

Trademarks. Suno, Udio, ACE-Step, DALL-E, Midjourney, Gemini, and Flux are trademarks of their respective owners. Lyricoro is not affiliated with, endorsed by, or sponsored by any of these companies. Product names are used solely to describe compatibility.

Contact

Question, suggestion, bug, or just want to say hi? Send a message — it goes straight to my inbox.

Message sent — thanks!