# Lyria 3 Pro - Core.Today AI API

> Google's most capable music generation model. Generates full-length songs up to 3 minutes with vocals, lyrics, song structure tags ([Verse], [Chorus], [Bridge]) and timestamp control from a text prompt, optionally inspired by up to 10 reference images. Returns an MP3 file.

- **Provider**: Google
- **Model ID**: google/lyria-3-pro
- **Category**: Audio & TTS
- **Credits**: 190 per song (up to 3 min)
- **Speed**: Medium
- **Quality**: Ultra

## Features
- Full-length songs up to 3 minutes
- Vocals, lyrics and song structure tags ([Verse], [Chorus], [Bridge])
- Timestamp control such as [0:00 - 0:30] for section timing
- Multi-language vocals
- Up to 10 reference images as mood inspiration

## Use Cases
- Original songs for videos, ads and branded content
- Demo tracks from written lyrics
- Theme songs and full-length background music

## API Endpoint

Base URL: https://api.core.today/v1
Create Prediction: POST /predictions
Get Status: GET /predictions/{job_id}

## Authentication
Header: X-API-Key: YOUR_API_KEY

## Input Parameters

### Required
- **prompt**: string - Text prompt describing the song. Include genre, instruments, mood, tempo, lyrics, and song structure (e.g. [Verse], [Chorus], [Bridge]). Timestamps like [0:00 - 0:30] control timing.

### Optional
- **images**: array (default: []) - Input images to inspire the music composition (up to 10 images)
- **seed**: integer - Random seed for more reproducible results (exact determinism not guaranteed)

## Examples

### Core.Today company song
The same Korean company-song lyrics given to every music model — compare how each one sings them
```json
{
  "model": "google/lyria-3-pro",
  "input": {
    "prompt": "[Verse]\n아침부터 AI, 커피보다 진해\n코딩하다 보니 벌써 새벽 두 시네!\n안 되는 건 없어, 일단 만들어 봐\n상상이 현실 되는 마법 같은 회사!\n\n[Chorus]\n코어! 닷! 투데이!\n오늘의 상상이 내일의 기술로!\n코어! 닷! 투데이!\nAI도 놀라는 우리의 아이디어!\n\n[Outro]\n뚝딱뚝딱 만들고, 반짝반짝 빛내고!\n기술을 넘어 경험으로!\n코어닷투데이, Hey!\n\nBright upbeat K-pop electro-pop company jingle, 128 BPM, energetic female and male vocals in Korean, catchy chant chorus, synth brass, punchy drums, about ninety seconds."
  }
}
```

### Song from structured lyrics
Tagged verse/chorus lyrics plus a style line become a full song
```json
{
  "model": "google/lyria-3-pro",
  "input": {
    "prompt": "[Verse 1]\nWalking through the neon glow,\ncity lights reflect below,\nevery shadow tells a story,\nevery corner, fading glory.\n\n[Chorus]\nWe are the echoes in the night,\nburning brighter than the light,\nhold on tight, don't let me go,\nwe are the echoes down below.\n\nSynth-pop, 110 BPM, female vocals, about ninety seconds."
  }
}
```

### Timed instrumental
Use timestamps to shape sections of an instrumental track
```json
{
  "model": "google/lyria-3-pro",
  "input": {
    "prompt": "Cinematic orchestral instrumental. [0:00 - 0:30] soft piano intro, [0:30 - 1:30] strings build with percussion, [1:30 - 2:00] full orchestra climax and gentle fade. No vocals."
  }
}
```

## Response Format
```json
{
  "job_id": "abc123",
  "status": "pending | processing | completed | failed",
  "result": "URL or data (when completed)"
}
```

## Usage Flow
1. POST /predictions with model and input -> receive job_id
2. GET /predictions/{job_id} -> poll until status is completed or failed
3. Result contains output URL(s)

## Tips
- Put lyrics in the prompt with [Verse]/[Chorus]/[Bridge] tags on their own lines; add a style line (genre, BPM, vocals)
- Say how long the song should be ('about two minutes') or use timestamps — otherwise length is up to the model
- Write 'No vocals' or 'Instrumental only' for instrumentals
- Reference images steer mood and theme — they also disable the fal fallback route
- For 30-second clips at half the price use Lyria 3; for the newest full-length model see Lyria 3.5