Mandarin city pop
Midnight Road
1:03Structured Chinese lyrics become a warm, compact city-pop song with an intimate female lead and a widened final chorus.
Custom lyricsFemale vocal108 BPM
Suno music model · March 2026
The whole song in one pass
Suno v5.5 AI Music Generator
Describe a song and Suno v5.5 writes the words, sings them, and arranges the parts around them. Or bring your own lyrics and direct the style line by line.
- v5.5
- Current model
- 2
- Takes per request
- 12
- Separable parts
- 30–120s
- Typical generation
Live composer
Make a song with Suno v5.5
Describe a complete song or bring structured lyrics. Two finished takes return from every request.
Suno v5.5
Includes both finished takes · Failed tasks are refunded
Your music
Private tracks stored in your Scenyra account
Your first song starts here
A detailed sentence is enough. Name the mood, instruments, tempo, and kind of voice you want to hear.
Generated with this page
Hear the prompt become a finished song.
These are real, unedited APIMart Suno outputs—not presenter videos. Play the full tracks, inspect the direction, then load the same brief into the composer.
Generated August 29, 2026 with Suno v5.5. Audio and cover art are served locally from Scenyra.
Cinematic instrumental
Dawn Over Ashes
1:58One prose brief moves from close felt piano to cellos, brushed percussion, and a restrained orchestral lift without a vocal.
InstrumentalOrchestralDynamic arc
Future gospel
We Rise
2:02A sparse male verse grows into layered choir, deep synth bass, cinematic trap drums, and a bright communal hook.
Custom lyricsMale vocalChoir + trap
Same prompt · two models
Compare v5 with v5.5 yourself.
Both tracks use the exact same Chinese lyrics, style tags, female vocal direction, and 108 BPM request. Only the requested model version changed. Because generation is stochastic, this is a practical listening comparison—not a laboratory benchmark.
Suno v5
Baseline output
Suno v5.5
Current output
The real generation process
From brief to the examples above.
- 01
Write the direction
We supplied mood, arrangement, voice, tempo, and either prose or sectioned lyrics.
- 02
Generate two takes
Each $0.05 APIMart request returned two independently composed full songs and covers.
- 03
Listen and choose
We ran two requests per concept, reviewed four candidates, and kept the clearest interpretation.
- 04
Publish the master
The selected MP3 and artwork were copied into local site assets; no expiring provider URL is used.
Showcase spend: 7 music requests × $0.05 ≈ $0.35 on APIMart. Fourteen candidates were generated; four audio files are published here.
Official walkthrough
See what changed in v5.5.
Suno introduces the model, demonstrates Voices, and shows how Custom Models bring a creator's own sound into future songs.
Read the official v5.5 announcementWhat v5.5 does
A complete song, then a production workflow around it.
Suno v5.5 can begin with one sentence or structured lyrics. The finished take becomes the source for later extensions, edits, stem separation, voice reuse, and exports.
Words, voice, and arrangement together
One request produces the lyric, vocal performance, arrangement, and finished song instead of stitching separate tools together.
Two takes, every time
Compare two full interpretations of the same brief and download either take from your private music history.
Vocal direction
Leave the singer open, choose a vocal gender, or render the idea as an instrumental with no vocals.
Twelve-part separation ready
Suno workflows can separate vocals, drums, bass, guitar, keys, percussion, strings, synth, FX, brass, and woodwinds.
Custom lyrics or pure inspiration
Let the model write from a one-line idea, or provide line-by-line lyrics with explicit song sections.
Permanent private files
Finished audio and artwork are moved to Scenyra storage immediately and served through an authenticated same-origin URL.
Voices
Keep a singer you created.
Suno launched v5.5 alongside Voices. A verified vocal identity from your own material can be saved privately and reused across future songs, helping a catalogue sound connected.
- 1Create a take with a vocal character you like.
- 2Verify and save the singer as a private Voice.
- 3Reuse that Voice when directing the next song.
Twelve stems
Hear the finished song split apart.
A full stem run can isolate the parts that matter to a mix, edit, game, podcast, or karaoke workflow.
Vocals
Backing vocals
Drums
Bass
Guitar
Keyboard
Percussion
Strings
Synth
FX
Brass
Woodwinds
How it works
From one idea to two finished takes.
Describe the song
Name the mood, instruments, tempo, and vocal direction—or write the exact lyrics and mark each section.
Compare both takes
Two complete songs return from the same request, giving you a useful creative choice immediately.
Keep and develop one
Download the master, then use its take identity for extensions, edits, voices, stems, WAV, MIDI, or video workflows.
Writing the prompt
Direct it like a producer.
Use style tags as constraints
Genre, mood, instrumentation, vocal character, tempo, and key work well as concise comma-separated tags in Custom lyrics mode.
Mark every lyric section
Put [Verse], [Chorus], [Bridge], and [Outro] on their own lines so the arrangement follows the intended shape.
Say what should be excluded
Use the advanced exclude field for unwanted genres, vocal techniques, or production traits instead of hiding them in prose.
Example
A slow orchestral dream pop song about leaving a city at dawn. Strings under a soft male lead, brushed drums, 96 BPM, warm tape hiss, building into a wide final chorus.
Questions
Suno v5.5 FAQ
Two complete takes