> For the complete documentation index, see [llms.txt](https://docs.qolaba.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.qolaba.ai/music-generation/generating-music.md).

# Generating Music

Music Generation in Qolaba is a full-featured Studio. This is where you configure every aspect of your generation — model, prompt, format, and output options — before generating.

***

### Step-by-Step: Generating Music

#### **Step 1 — Write a Prompt** 

Describe the music you want in the prompt field. Be specific about genre, mood, instruments, tempo, and any lyrical theme.

* **Minimum:** 2–3 words
* **Maximum:** 5,000 characters

**Strong prompt examples:**

```
Upbeat pop song about summer adventures with catchy
chorus, electric guitar, and synthesizers.
```

```
Ambient lo-fi beat for studying — soft piano, gentle
rain sounds, slow tempo, no vocals.
```

```
Cinematic orchestral piece — dramatic strings, building
tension, suitable for a film trailer.
```

#### **Step 2 — Select a Model** 

Choose from **Lyria 3 Clip and Lyria 3 Pro.** Each model is displayed with its credit cost. See [Models & Configuration → ](/model-reference/music-models.md) for a full breakdown of model capabilities and differences.

#### **Step 3 — Configure Options**

| Option                | What It Does                                                                                     |
| --------------------- | ------------------------------------------------------------------------------------------------ |
| **Instrumental Only** | Toggle on to generate music without vocals                                                       |
| **Format**            | Choose MP3 (smaller, ideal for sharing) or WAV (high-fidelity, for editing) where supported      |
| **Negative Prompt**   | Describe what to exclude — e.g., *"no drums, no distortion, no fast tempo"* (select models only) |
| **Seed**              | Enter a number for reproducible results — same seed + same prompt = same output every time       |

#### **Step 4 — Use Magic Prompt (Optional)** 

Click **Magic Prompt** to have the AI automatically enhance your prompt before generation. This expands a short or simple prompt into a more detailed, model-optimized version — producing richer, more specific musical output.

* **Cost:** \~0.05 credits per enhancement
* **Best for:** Short or underdeveloped prompts where you want more stylistic detail without writing it manually

#### **Step 5 — Generate** 

Click **Generate**. Credits are deducted upon completion.

***

#### Writing Effective Music Prompts

<table><thead><tr><th width="211.76171875">Element</th><th>What to Include</th></tr></thead><tbody><tr><td><strong>Genre</strong></td><td>Pop, rock, jazz, classical, lo-fi, ambient, cinematic, electronic</td></tr><tr><td><strong>Mood</strong></td><td>Upbeat, melancholic, dramatic, calm, energetic, mysterious</td></tr><tr><td><strong>Instruments</strong></td><td>Guitar, piano, strings, synthesizers, drums, bass</td></tr><tr><td><strong>Tempo</strong></td><td>Slow, mid-tempo, fast, building</td></tr><tr><td><strong>Vocals</strong></td><td>With or without vocals, male/female voice, choir</td></tr><tr><td><strong>Theme</strong></td><td>Subject matter for lyrics if vocals are included</td></tr></tbody></table>