Suno — Complete Songs, From One Sentence or a Full Lyric Sheet
Two Ways In: Describe a Song, or Direct One
Suno generates complete music from text, and every request returns multiple variations of your song. In simple mode you write one description — like the documented example, 'A cheerful summer pop song about road trips and freedom' — and the AI generates the lyrics and style for you. In custom mode you take the producer's chair: define the style in comma-separated tags, name the track, supply exact lyrics with [Verse]/[Chorus]/[Bridge] structure tags, exclude styles you don't want, and state a vocal gender preference. Five model versions are available, from V4 up to the latest V5; V4.5 and above generate tracks up to 8 minutes long.
The two modes map to two different creative situations. Simple mode takes a single description of up to 500 characters and handles everything else — lyric writing and style selection included. Custom mode unlocks the full input surface: a style field (up to 1,000 characters on V4.5+, 200 on V4) using comma-separated tags for genre, mood, instruments, vocals, and tempo; a title of up to 80 characters that appears in the player and filename; and your prompt becomes the exact lyrics, up to 5,000 characters on V4.5+ (3,000 on V4), organized with structure tags like [Verse], [Chorus], and [Bridge]. An instrumental switch produces vocal-free music — in custom mode, lyrics become optional when it's on. Negative tags exclude styles ('heavy metal, screaming, sad'), and the vocal gender option nudges toward a male or female voice — per the documentation it raises the probability rather than guaranteeing it.
Version choice matters: V5 is the latest with superior musical expression and faster generation; V4.5+, V4.5 All, and V4.5 generate up to 8 minutes with richer tones and smarter prompts; V4 offers improved vocal quality with tracks up to 4 minutes and tighter character limits. Every run returns multiple variations, so each generation gives you alternatives to choose from. Generation is asynchronous — progress shows on the page, and finished tracks are saved to your dashboard gallery.
How to Generate a Song with Suno
Pick a Version & Mode
Choose a model version (V5 is the latest), then decide: simple mode for a one-line description, or custom mode for full control.
Write the Inputs
Simple mode: describe the song in up to 500 characters. Custom mode: set style tags and a title, paste lyrics with [Verse]/[Chorus] tags, optionally exclude styles and set a vocal gender preference.
Generate & Pick a Variation
Each run returns multiple variations — preview them in the players, download your favorite, and find everything again in your dashboard gallery.
Suno Technical Specifications
| Provider | Suno (via Evolink API) |
| Model Versions | V5, V4.5+, V4.5 All, V4.5, V4 |
| Simple Mode Prompt | Up to 500 characters — AI writes lyrics & style |
| Custom Mode Style | Up to 1,000 characters (V4: 200) |
| Custom Mode Lyrics | Up to 5,000 characters (V4: 3,000), structure tags supported |
| Title | Up to 80 characters |
| Max Track Length | 8 minutes (V4.5+), 4 minutes (V4) |
| Instrumental Mode | Yes |
| Vocal Gender Preference | Male / female (raises probability, not guaranteed) |
| Variations per Run | Multiple |
| Processing | Asynchronous (queue-based) |
Why Choose Suno
Multiple Variations Every Run
One generation returns several takes on your song, so you choose the best version instead of settling for a single roll.
Beginner to Producer
Simple mode turns one sentence into a finished song; custom mode exposes style tags, titles, exact lyrics, negative tags, and vocal gender preference.
Five Versions, Up to 8 Minutes
Choose from V4 through the latest V5 — V4.5 and above generate tracks up to 8 minutes with richer tones and smarter prompt handling.
Suno vs Other Music Models on LoveGen
| Feature | Suno | MiniMax Music 2.6 | ElevenLabs Music |
|---|---|---|---|
| Input Modes | Simple (description) or custom (style + title + lyrics) | Style prompt, lyrics optional | Text prompt |
| Auto-Written Lyrics | Yes — simple mode | Yes — lyrics optimizer | No |
| Instrumental Mode | Yes | Yes (is_instrumental) | Yes — guaranteed toggle |
| Vocal Gender Preference | Yes (m/f, probabilistic) | — | — |
| Max Track Length | 8 min (V4.5+), 4 min (V4) | Model decides | Exact, 3 seconds–10 minutes |
| Variations per Run | Multiple | 1 | 1 |
What Can You Create with Suno
Songs from a Sentence
Simple mode turns a one-line idea into a complete song with AI-written lyrics and style — the fastest way from concept to track.
Original Songs from Your Lyrics
Custom mode performs your exact words, structured with [Verse]/[Chorus]/[Bridge] tags, in the style you specify.
A/B Picking
Multiple variations per run mean every generation is a mini listening session — keep the take that lands.
Instrumental & Background Music
Flip the instrumental switch for vocal-free tracks driven by your style tags alone.
Long-Form Tracks
V4.5 and above generate up to 8 minutes — room for full song structures beyond the radio edit.
Style-Controlled Production
Steer with comma-separated style tags, exclude what you don't want via negative tags, and nudge the vocal gender.
Related AI Models
MiniMax Music 2.6
Complete tracks with singing & detailed arrangements — auto lyrics & instrumental mode
ElevenLabs Music
Studio-grade AI music from a text prompt — vocals or instrumental, up to 10 minutes
Sonilo v1.1
Production-ready music from one prompt — exact duration control, up to 10 minutes
CassetteAI
Lightning-fast music generation — a full 3-minute track in seconds
Frequently Asked Questions About Suno
What's the difference between simple and custom mode?
Simple mode takes one description (up to 500 characters) and the AI generates the lyrics and style automatically. Custom mode gives you direct control: style tags and a title are required, your prompt becomes the exact lyrics, and you can add negative tags and a vocal gender preference.
Which version should I pick?
V5 is the latest, with superior musical expression and faster generation. V4.5+, V4.5 All, and V4.5 generate up to 8 minutes with richer tones. V4 offers improved vocal quality, up to 4 minutes, with smaller input limits (3,000-character lyrics, 200-character style). When unsure, start with V5.
How many songs do I get per generation?
Each request generates multiple variations of your song. All of them appear in the output players when the task completes, and all are saved to your dashboard gallery.
Can I control the voice?
In custom mode you can set a vocal gender preference (male or female). Per the API documentation, this raises the probability of getting that voice but doesn't guarantee it. For music without vocals entirely, enable instrumental mode.
How long does generation take?
Generation runs on an asynchronous queue, so the wait varies with version and current load. Progress is shown on the page, and finished tracks are also saved to your dashboard gallery, so you can navigate away safely.
How much does a generation cost?
Each generation costs a flat number of credits regardless of which variation you keep — the exact amount is shown on the Generate button before you submit. Failed generations are refunded automatically. Download tracks within 7 days; gallery media is cleaned up after that window.