Overview
Turn a hummed melody into a vocal song draft. Provide an audio reference, add lyrics, describe the sound in Style, and choose the requested song duration. The configured output is WAV. Listen for how the melody and words come through, including phrasing and section starts, before deciding what to revise.
AI Hum to Song uses a supplied melody idea with your lyrics and Style description to generate a vocal song. Duration sets the requested length of the generated song. Listen to the result to check how the melody and words were interpreted.
When to use
- Use AI Hum to Song when you have a melody idea and want to develop it into a vocal song draft with your lyrics and a Style description.
- If you want the title or artist of a tune you remember, use a music-identification service instead. AI Hum to Song generates music from a supplied idea; it does not search for an existing song.
How to use
Open AI Hum to Song in its linked workspace. This online AI workflow requires sign-in and an internet connection.
Add one melody recording in .mp3, .wav, .ogg, .flac, .m4a, or .aac format. The file can be up to 10 MB and 30 seconds long. Record or select a clip where you can clearly hear the melody line you want to develop. The listed extensions alone don't confirm codec or bit-depth support, so check the workspace's visible restrictions before continuing. Enter your words in Lyrics and describe the sound you want in Style. In Lyrics, you can use [verse], [chorus], and [bridge] tags; keep lines short, around 6–12 syllables. For example, Style could say “Acoustic folk, gentle vocal, fingerpicked guitar.”
Choose Duration: 30 Seconds, 60 Seconds, or 120 Seconds. This sets the requested length of the generated song, not the length of your recording. Review the displayed charge, then use the workspace action to process. Listen for how the melody was interpreted, whether the words are pronounced as intended, and where the sections begin. If a WAV download is offered, check the downloaded file before using it.
Settings reference
| Setting | Options | Default |
|---|---|---|
| Style | Text | Empty |
| Lyrics | Text | Empty |
| Duration | 30: 30 Seconds 60: 60 Seconds 120: 120 Seconds | 60 |
Formats & limits
| Detail | Requirement |
|---|---|
| Input formats | .mp3, .wav, .ogg, .flac, .m4a, .aac |
| Output formats | WAV |
Access & cost
| Detail | Requirement |
|---|---|
| Processing mode | Online processing requires sign-in and an internet connection. |
| Online processing | Sign-in required; Internet connection required |
Examples & use cases
A songwriter developing a hummed hook
Develop a hummed hook
They have a .wav melody idea and want to hear it developed into an acoustic folk draft. Their lyrics have a verse and a short chorus, and a one-minute version gives them room to review both sections.
.wav to WAV
- Style:
- Acoustic folk, gentle vocal, fingerpicked guitar
- Lyrics:
- [verse] Rain taps softly on the glass [chorus] Stay here until the morning
- Duration:
- 60
- Choose 60 Seconds to request a one-minute song draft with room to review the hook and both lyric sections.
- Describe the acoustic folk sound in Style. Mark the verse and chorus in Lyrics so you can check how the generated song places them.
- Listen to the WAV and check the hook, lyric pronunciation, and section starts.
Expected outcome
The configured output is WAV with a requested song length of 60 seconds. Listen for whether the hook is recognizable, the words are pronounced as intended, and the verse and chorus start where you want them.
A creator sketching a short video theme
Sketch a short theme
They have an .mp3 melody idea and want a brief vocal draft for a short video. A 30-second requested song length keeps the draft focused on the theme rather than extending it into a longer song.
.mp3 to WAV
- Style:
- Bright pop, upbeat vocal, light synths
- Lyrics:
- [verse] A little spark across the sky [chorus] We rise and let the colors fly
- Duration:
- 30
- Choose 30 Seconds for a brief generated-song draft; this setting doesn't set the length of the uploaded reference.
- Use Style to describe the pop sound and vocal character. Add the words in Lyrics and mark sections if you have a particular structure in mind.
- Listen to the WAV and check how the tune and words come through in the short draft.
Expected outcome
The configured output is WAV with a requested song length of 30 seconds. Listen through the draft to check how the melody and words come through and whether the section change suits the intended theme.
A lyricist reviewing a longer song idea
Review a longer lyric draft
They have an .m4a melody reference and lyrics planned across a verse, chorus, and bridge. They want a longer draft so they can listen to how the words and sections work together.
.m4a to WAV
- Style:
- Soft rock, expressive vocal, piano and drums
- Lyrics:
- [verse] The road is quiet after rain [chorus] We find our way through every change [bridge] The lights are waking in the town
- Duration:
- 120
- Choose 120 Seconds to request a longer song draft for reviewing the planned sections; this doesn't change the permitted length of the melody reference.
- Mark the planned sections in Lyrics and describe the vocal character and instrumentation in Style.
- Listen to the WAV and check how the words are phrased and whether the section starts fit the lyric structure.
Expected outcome
The configured output is WAV with a requested song length of 120 seconds. Listen for phrasing, pronunciation, and where the verse, chorus, and bridge begin before deciding what to revise.
More ways to use it
- Songwriters exploring a melody: A songwriter has a melody idea and wants to hear it developed with lyrics and a described musical style, rather than use the melody only as a recording. A vocal song draft with a requested length of 60 seconds and a configured WAV output, ready to listen to as a starting point for further writing. Style: A description of the sound they want, Lyrics: Lyrics for the song, Duration: 60.
- Lyricists checking words against a melody: A lyricist wants to hear how their words are phrased with a melody reference and where the generated song places its marked sections. A song draft to review for phrasing, pronunciation, and section starts. Listen to the result to decide whether the lyrics or Style description need revision. Style: A description of the sound they want, Lyrics: [verse] ... [chorus] ..., Duration: 60.
Troubleshooting
The song structure doesn't come through as intended.
Try this: In Lyrics, mark the intended sections with [verse], [chorus], or [bridge] tags.
What to expect: The tags mark the intended song sections. Listen to the result to check how the generated song places them.
The requested song is shorter or longer than the draft you want to review.
Try this: Choose the Duration value that matches the requested generated-song length: 30, 60, or 120.
What to expect: The selected Duration sets the requested song length; 30 requests 30 seconds. It doesn't change the length of the melody recording.
Privacy
Online processing uploads files. Completed results may remain available for later download. See Privacy Policy for current file-handling and retention details. Check the processing mode before starting.
Frequently asked questions
Can I leave Lyrics blank?
No. This workflow requires lyrics as well as a Style description and an audio reference. Add your words in Lyrics; you can mark sections with [verse], [chorus], and [bridge] tags.
Can I upload a two-minute melody recording?
No. A melody recording can be up to 30 seconds long. The 120 Seconds option sets the requested length of the generated song, not the length of your recording.
What is the difference between humming a song idea and identifying an existing song?
AI Hum to Song generates music from a supplied melody idea. It doesn't search for the title or artist of a tune you remember; use a music-identification service for that.
Which audio formats can I upload, and what does the tool produce?
The declared input extensions are .mp3, .wav, .ogg, .flac, .m4a, and .aac. The configured output is WAV. The listed extensions alone don't establish codec or bit-depth support.
How long can the generated song be?
Choose 30 Seconds, 60 Seconds, or 120 Seconds in Duration. This sets the requested length of the generated song, not the length of your uploaded melody recording.
Can I choose a genre, mood, or instruments?
Use Style to describe the sound you want, such as “Acoustic folk, gentle vocal, fingerpicked guitar.” Listen to the result to check how it interpreted your description.
Can I add my own lyrics and mark song sections?
Yes. Enter your words in Lyrics and use [verse], [chorus], and [bridge] tags to mark sections. The documented guidance is to keep lines short, around 6–12 syllables. Listen to check how the words are phrased and where sections begin.
Will the generated song reproduce my melody exactly?
The tool generates a song from your supplied melody idea, lyrics, and Style description; it isn't a promise of note-for-note reproduction. Listen to the WAV to check how the melody was interpreted, including its phrasing and section starts, and decide whether to revise the words or Style.
Can I download the generated song?
The configured output is WAV. If the workspace offers a download, check the WAV file and listen for how the melody, pronunciation, and section starts came through.
Is anything uploaded during processing?
Online processing uploads files. Completed results may remain available for later download. See Privacy Policy for current file-handling and retention details. Check the processing mode before starting.
Does this workflow work offline?
No. This online AI workflow requires an internet connection and sign-in to process the melody recording.
Can I upload a voice memo instead of recording humming in the browser?
You can use an audio reference file in .mp3, .wav, .ogg, .flac, .m4a, or .aac format. The available product facts don't specify a browser recording control, so use a supported audio file as the melody reference.
How do I turn a melody I hum into a full song?
Provide an audio file containing the melody, add your lyrics in Lyrics, describe the sound in Style, and choose 30, 60, or 120 Seconds in Duration. Process the draft, then listen to how the melody, words, and marked sections came through.