Suno V5.5 Guide: Create Your Own Voice in Suno (Step-by-Step)

Suno recently released v5.5, bringing 3 revolutionary new features: Voices, Custom Models, and My Taste.

As a user who's been on Suno since v4, I spent the first week after the v5.5 launch thoroughly testing all the new features. In this Suno v5.5 tutorial, I'll walk you through these 3 core features step-by-step, helping you go from beginner to expert.

Let's dive in!

Part 1. 3 Core New Features Explained

1 Voices: Use Your Own Voice in AI Songs

Voices is Suno's most requested feature, allowing you to use your actual voice to sing AI songs. This isn't simple voice cloning, it captures your voice characteristics like timbre, texture, and range, then integrates these qualities into Suno's music generation.

Note that Voices is only available for Pro or Premier subscribers, and you can create multiple Voice profiles.

How to Use Voices Feature?

Step 1Open the Suno Create page and click the Voice button (formerly the Personas button) or the Try Now prompt.

Step 2Click Create Voice in the popup panel to enter the creation window.

Step 3Provide a voice sample. You have 3 options: record or upload your voice or select from library. Recording requires a minimum of 10 seconds.

Step 4After checking the authorization, you'll need voice verification. Suno displays a random phrase for you to read aloud and record.

Step 5Click Save to save your Voice profile and give it a name. I recommend naming it based on its intended use.

Step 6Return to the Create page, click Voice and select your newly created Voice from the top of the list. You can adjust Audio Influence to set how much your voice affects the output.

Step 7Click Create, and Suno will generate two songs using your Voice.

• My Actual Test Results
• I created a Voice profile with my own voice and tested different styles. Each song carried my vocal characteristics and folk sounded closest to my real voice, but with much better singing skills than I actually have.

2 Custom Models: Train AI on Your Music Style

Custom Models lets you upload your original music works to train your personalized Suno v5.5. The trained model will mimic and learn your music style. You can create up to 3 custom models, and each training session requires uploading at least 6 songs that you own the copyright to.

Custom Models vs Voices

Voices: Only affects vocal characteristics.

Custom Models: Affects overall music style.

How to Use Custom Models Feature?

Step 1Prepare your training material. Before creating a Custom Model, you need at least 6 original songs you've created. These songs should maintain style consistency. if the training material is too scattered, AI can't extract clear style characteristics.

I recommend creating different Custom Models for different frequently-used styles.

Step 2Open the Suno Create page, click the model in the upper right, and choose Create Custom Model from the dropdown menu. This opens the Custom Model creation interface where you can name your model.

Step 3Upload your music. You can upload individually or in batches.

Step 4Once you've uploaded at least 6 songs, click Create Custom Model to create and train the model.

Step 5After training completes, your Custom Model appears in the model menu. Select it and input your prompt to use.

• My Actual Test Results
• I created a Custom Model based on 8 lo-fi hip-hop tracks I composed. Then I used both v5.5 and my Custom Model with the same prompt. The results were clear, songs generated with my Custom Model better matched my style and taste.

3 My Taste: Personalized Recommendations

My Taste is a fully automated personalization system. It learns from the songs you create and listen to on Suno, then automatically provides style suggestions that match your taste when you create new music.

This feature is available to all users, enabled by default, and requires no setup or training.

How to Use My Taste Feature?

Step 1Go to the Suno Create page and select Advanced mode.

Step 2Next to the Style input box, there's a magic wand icon. Click it, and Suno will automatically generate a personalized style prompt based on your My Taste profile.

• My Actual Test Results
• I'm a huge fan of lo-fi hip-hop, and previously Suno's style suggestions didn't quite match my creative habits. But after using Suno v5.5 for two weeks, the generated style suggestions became noticeably more accurate. Now when I lack inspiration, I always use it.

How to Manage My Taste Feature?

My Taste is enabled by default, but you can control it.

Step 1Click your avatar and select My Taste.

Step 2In the popup interface, you'll see Suno's summary of your music taste profile. You can manually edit it or click the disable button below.

Part 2. Practical Tips for Using Suno v5.5

After mastering 3 major features, these practical tips will take your Suno v5.5 experience to the next level.

1 Provide High-Quality Voice Samples

The Voices feature's performance heavily depends on your audio sample quality. I recommend maintaining consistent tone when recording, capturing over 20 seconds of clear audio, and covering different pitch ranges and emotional variations.

2 Adjust Audio Influence Reasonably

Through extensive testing, I've found that setting Audio Influence between 70%-90% usually works best, preserving your voice characteristics while ensuring the overall music sounds natural.

3 Keep Style Consistent Within Each Custom Model

The core of Custom Models is learning style, not individual songs. If your uploaded training data has mixed styles, the model output becomes unstable. I recommend creating separate models for each style to achieve more consistent results.

4 Try Using Negative Prompts

Suno v5.5 supports negative prompts, you can explicitly tell the AI what you don't want, effectively excluding unwanted elements from your songs.

5 Combine Voice and Custom Models for Personalized Music

Like traditional music production, you can flexibly use different voices and styles, creating multiple combinations.

6 Use My Taste to Break Through Creative Blocks

When you're stuck on what to create, try clicking the magic wand to see what style suggestions Suno generates based on your music taste profile. Overall, I find this is an excellent creative starting point.

Part 3. Suno v5.5 Strengths & Limitations (Honest Review)

Here's my honest assessment after using Suno v5.5 for nearly two weeks.

1 Strengths

1. Voices and Custom Models Enhance Personalization

This is v5.5's biggest breakthrough. Through Voices and Custom Models, Suno has moved from generic generation to personalized creation. You can make AI mimic your voice or train specific styles, making generated music more distinctive.

I tested music generated by a friend using his Custom Model versus mine, the results were completely different.

2. Voices Lowers the Barrier for Vocal Demos

For producers who aren't professional singers, the Voices feature is a godsend. Previously, I needed to find singers to record demos or record myself. Now I can quickly demo melodies and lyrics with my voice. Even if my singing isn't great, AI beautifies it while preserving my vocal characteristics.

3. Stable Overall Generation Quality, Suitable for Batch Production

For most common styles, Suno's output quality is already good enough for short video soundtracks or content creation. I specifically created a Custom Model for video BGM. When I need batch production, I can use it directly for decent results.

2 Limitations

1. Voices are Unstable and Unpredictable

In actual Voices usage, I've encountered the same issues other users report, such as voices suddenly going higher, adding unnecessary runs, or sometimes sounding nothing like me, becoming exaggerated and distorted.

2. Generated Songs Have Noticeable Audio Quality Issues

Some songs I generated with Suno v5.5 have noise or hissing that makes them completely unusable. This issue is widespread and even affects songs generated with Suno v5.

3. Unstable Full-Song Structure

Through extensive testing, I've found many songs that sound great at the beginning suddenly fall apart at the end. There are also cases where the verses are excellent but the chorus deviates from the song's style. This often requires repeated generation, wasting significant time.

• Many people wonder whether to subscribe to Suno v5.5. My recommendation is if you want to create personalized AI music and are willing to invest time learning and training, v5.5 is absolutely worth it.
• But if you're on a tight budget or just want to quickly generate music, Suno's free version or other tools might be simpler and more straightforward.

Part 4. Easier Alternative: Create Music with Tunesona

After using Suno v5.5 extensively, I want to introduce a music agent worth noting—Tunesona. It's not a Suno clone but offers a completely different music creation experience. If Suno v5.5's breakthrough is personalization, then Tunesona's core advantage is conversational creation.

You don't need to learn complex prompts or pre-train models, just describe your ideas like chatting with a music producer, and AI understands and creates music matching your intent.

More importantly, Tunesona supports multi-turn iterative dialogue. Whether it's melody, rhythm, or lyrics, you can adjust and customize based on your preferences. This workflow is closer to the real music creation process.

Furthermore, if you still prefer more granular control, Tunesona also offers a custom mode, perfectly suited for users of different skill levels.

You can use Tunesona to:

• Generate lyrics. For example, when you enter "I want to generate lyrics for my daughter's birthday," Tunesona will provide 3 different versions of the lyrics.
• Recommended styles. For example, when you enter "I want some rock music," Tunesona will recommend 3-6 music styles to inspire you.
• Generate music. Tunesona can generate instrumental music and music with lyrics based on your description.
• Audio references. You can upload audio as style, melody, or vocal references.
• Freely expand audio. You can freely choose where and how to expand uploaded or generated audio.
• Automatic mixing. Tunesona can automatically balance different parts, add instrument tracks, etc., based on your description.

Suno v5.5 vs Tunesona Comparison Table

Part 5. FAQs About Suno v5.5 Tutorial

1 How long does it take to train a Custom Model?

Typically 2-5 minutes, depending on the number of songs uploaded, song length, and complexity.

2 Can I use Voices with someone else's voice?

No. Suno has a strict identity verification process. You must read a random phrase for voice verification, the system compares your uploaded vocals with the verification recording to confirm they're the same person before creating the Voice.

3 What's the difference between Voices and Personas?

Voices is an upgraded version of Personas. Voices is more realistic and controllable, while Personas was an earlier, more basic vocal preset feature.

4 Can I share my Custom Models with others?

No. Custom Models are private and only you can use them. This protects copyright and your privacy.

Conclusion

Suno v5.5 isn't just a faster, stronger upgrade, it's a personalization evolution. If you want to try these new features, you can follow the steps in this Suno v5.5 tutorial.

But that doesn't mean Suno v5.5 is right for everyone. If you prefer conversational creation and quick generation, Tunesona might be a better choice.