Suno Persona API Guide: Achieving Consistent Vocal Style

TL;DR

• Goal: Maintain vocal and style consistency across different music generations.
• Key Parameters: persona_id and custom_mode=True.
• Workflow: Create a Persona from an existing track → Retrieve persona_id → Use in subsequent generations.
• Integration: Pass persona_id during the /v1/audios/generations task submission.

What is Suno Persona API?

Why Use Persona API?

• Branding: Maintain a consistent "brand voice" for podcast intros or ad jingles.
• Consistency: Keep the same vocalist for episodic series or multi-part narrative songs.
• Character Design: Assign specific vocal "personas" to characters in a game or video project.

How to Get Started with Persona API

1. Generate a Baseline Track: Create a Suno track using the standard generation API.
2. Create Persona: Call the Persona creation endpoint using the id of your successfully generated track.
3. Retrieve ID: The API will return a unique persona_id.
4. Reuse: Pass this persona_id in future generation requests to apply the vocal style.

Technical Implementation

Submit Generation Task with Persona

import requests

API_KEY = "sk-your-evolink-key"
BASE_URL = "https://api.evolink.ai/v1"

headers = {
    "Authorization": f"Bearer {API_KEY}",
    "Content-Type": "application/json"
}

def submit_persona_task(prompt, persona_id):
    url = f"{BASE_URL}/audios/generations"
    payload = {
        "model": "suno-v5-beta",
        "prompt": prompt,
        "persona_id": persona_id,
        "custom_mode": True
    }
    response = requests.post(url, headers=headers, json=payload)
    response.raise_for_status()
    # Returns 'id' as the task identifier
    return response.json()["id"]

task_id = submit_persona_task(
    prompt="Upbeat pop chorus about city lights", 
    persona_id="your-unique-persona-id"
)
print(f"Task submitted with id: {task_id}")

Retrieving Results

def get_song_result(task_id):
    url = f"{BASE_URL}/tasks/{task_id}"
    while True:
        response = requests.get(url, headers=headers)
        data = response.json()
        
        if data["status"] == "completed":
            # Access result from result_data array
            return data["result_data"][0]["audio_url"]
        elif data["status"] == "failed":
            raise Exception(f"Task failed: {data.get('error')}")
        
        time.sleep(5)

Best Practices

• Style vs. Content: persona_id locks the vocal characteristics. Ensure your prompt style is compatible with the persona’s baseline.
• Custom Mode: Always set custom_mode=True when using persona parameters, as this is a documented requirement for the API to process them correctly.

Troubleshooting & FAQ

Resources

• Suno API Overview
• Suno API Tutorial
• Music Generation Documentation
• Persona Feature Documentation