AI MODEL DISCOVERY
TypeSafeTypeSafe Jev 1.13
Structured decisions from text or JSON, with yes/no probabilities, choices and scores.
Start with the words you want to sing. Use your lyrics, a song title and a musical style to create vocal songs with AI Music M1, M2 or M3 through SeeAPI's API.
PROVIDERS
Build a songwriting flow that starts with the words rather than a scene description.
Send your lyrics with a title and style in custom mode. Choose vocal output to turn the written lines into a song.
Alibaba
Black Forest Labs
ByteDance
Google
Grok
Kling
Lightricks
MiniMax
OpenAI
PixVerse
Runway
Vidu{
"model": "ai-music-m2",
"endpoint": "music-generation",
"input": {
"custom_mode": true,
"instrumental": false,
"title": "A New Day",
"style": "Warm acoustic pop, gentle vocals",
"lyrics": "The morning opens like a door\nA little brighter than before"
}
}Describe the genre, instruments and vocal feel in style, then compare the arrangements returned for your lyrics.
Read task status and stage updates. Preview available audio while processing, then retrieve the completed tracks.
GET /v1/generations/{id}
processing → text
processing → first
succeeded → completeEach generation includes all tracks returned for that task, with audio, lyrics and cover artwork when available.
Create a task with your own lyrics and collect its sung versions.
Compare AI Music M1, M2 and M3 with the same lyric draft and style. Check the documented lyrics limit and current API price, then judge phrasing, vocal delivery and arrangement against the song you want to make.
Keep sung words in lyrics, with line breaks and optional [Verse] or [Chorus] labels. Put the song name in title and musical direction in style. Use custom_mode: true and instrumental: false for a vocal song.
Poll the task or receive callbacks, and preview when a playback URL is returned. Once status is succeeded, collect every track in result.assets. Review the sung result and keep each track's ID with its audio and returned lyrics.
Give lyric writers a way to hear and compare musical interpretations.
Turn a lyric draft into vocal demos so writers can listen to how their words work with different musical directions.
Let users add music to a poem or character's song by supplying lyrics and a fitting style.
Compare sung versions of a lyric exercise and explore how musical style changes the listening experience.
Build a flow where users supply words for a birthday, a reunion or a personal story. Pair their lyrics with a title and musical style, then let them review the songs.
Develop a short hook around a campaign message. Keep the words stable while comparing musical directions, then review pronunciation, pacing and how clearly the message comes through.
Bring generated demos into your team's writing workflow. Keep lyric versions and task IDs together so collaborators can discuss which lines, phrasing and arrangements to refine next.
Practical details for choosing a model and integrating music generation.
Use custom mode with instrumental set to false. Supply your lyrics, a title and a style, then retrieve the tracks when the task succeeds. See the model documentation for required fields and limits.
Open the API reference for model parameters and response examples, or compare the current credit price for each model.