ACE-Step 1.5 and Mureka V8 Now Live! Review & Win Free Membership🔥

Announcements
1

Dear Tuneees,

We are thrilled to announce that ACE-Step v1.5 and Mureka V8 have officially landed on Tunee. This update marks a dual breakthrough in generation efficiency and output quality, empowering creators to produce high-fidelity AI music faster and more cost-effectively than ever.

:new_button: What’s New in ACE-Step v1.5?

  • Studio-Grade Quality: Delivers commercial-standard audio suitable for professional projects and mainstream media.
  • Ultra-Fast Inference: Significantly reduced generation times with optimized prompt recognition.
  • Limitless Versatility: Supports 1,000+ instruments and genres, responding accurately to nuanced tonal commands.
  • Global Language Support: Compatible with 50+ languages; use lyrics to dictate song structure and stylistic shifts.
  • Full-Length Compositions: Generate complete tracks up to 10 minutes long.

A Note on Performance: While we are pleased to offer this expanded range of languages, please be aware that performance may vary. Due to differences in training data volume across languages, the naturalness of certain regional dialects or less common languages is still being evaluated. We welcome your honest feedback to help us refine these capabilities in future updates.

:flexed_biceps: Mureka V8: A Major Leap in Professional Music Creation

  • Polished, complete tracks: Catchy melodies with tight structure, clear verse-chorus contrast, no more fragmented audio — just fully realized, release-ready songs.
  • Ultra-realistic vocals: Tone and delivery are precisely matched to the lyrics, with natural flow and rich emotion that captures the feel of a real lead vocal.
  • Studio-grade arrangements: Rich, layered instrumentation with clean, well-structured harmonies that deliver powerful, professional sound.
  • Musical intelligence: Understands song structure, harmonic logic, and emotional arc — resulting in cohesive, fluid compositions with a complete narrative feel.

:rocket: Our Model Matrix: Staying Current

Our mission remains unchanged: To integrate the most capable models as they emerge, ensuring our users always have access to the latest creative technology.

  • Last Week: We successfully deployed MiniMax 2.5, enhancing text comprehension and emotional depth in generation.

  • Coming Soon: We are in the final stages of debugging Mureka V8, a flagship model that will be available to you very soon.

:guitar: Feature Update: Cover Function

We are also prioritizing the restoration of the Cover (Song/Lyric Replacement) feature. Our team is currently adapting this function to our new model architecture. It will return shortly, designed to be more intuitive and responsive to your creative needs.

:heart: A Final Word

Tunee is a young team. While we have faced challenges along our journey, we have never considered giving up on our users. Thank you for staying with us and for your continued patience and feedback.

We invite you to log in to Tunee and resume your creative journey today.

— The Tunee Team

Community Event: Share Your Feedback

Try the new models and share your test results and user experience in the comments!

We will select 3 users from the comment section to receive one month of Tunee Plus for free.

:trophy: Winners will be announced on March 1st.

:tada: Update: Winners Announced!

We have selected three users from the comments section and granted them a one-month Tunee Plus membership as a reward.

Please kindly check your email inbox and direct messages for further details.

Once again, thank you to everyone who participated and shared your valuable feedback — we truly appreciate your support!

3 Likes
4

Actually I tried to use it for Farsi (Persian) and it was useless. Very sad. But the more important thing is this reverberation or echo at the end of a long song. That makes it almost impossible to use this programm, for that I pay. Why I can not contact a real person - a real support? I want to get a part of my money back.

6

Ten model Ace-step 1.5 jest również fatalny jak poprzednie z tego roku ,a piosenki w języku polskim to porażka.

7

I‘ve tried it. Unfortunately, it can’t sing the name Joséphine correctly in French.

8

thank you for bringing back the support for Vietnamese language

I have tested and tested and all the tracks sound like some foreign people singing in Vietnamese. Therefore, ACE-step 1.5 is unusable

But the beat and flow are fresh and smart, which is a good thing.

Thanks again, and I am faithfully awaiting for your future improvements

regards,

SP

9

Hello, we’ve checked your billing information in our system and found that the payment was made on January 25, 2026. As it has now been more than 7 days since the purchase, it falls outside our refund policy period.

If you have any further questions or concerns, please feel free to contact us at support@tunee.ai.

10

Hi, just send you my request again. Since January this is happening. Can you please answer why your program produces echoes and halls with long songs. I just tried it yesterday again. These songs are useless.

Kind regards, Michael L.

y I tried to use it for Farsi (Persian) and it was useless. Very sad. But the more important thing is this reverberation or echo at the end of a long song. That makes it almost impossible to use this programm, for that I pay. Why I can not contact a real person - a real support? I want to get a part of my money back

1 Like
12

Tried ACE with 12 different songs. All (!) songs had distorted sound (sounds like a volume slider is set to a much to high level). And a lot of songs were disharmonic, the vocals had a wrong intonation and sounded as if the singer has no talent and is singing in a complete different tune ….

1 Like
13

We’re sorry to let you know that your request is outside of our refund eligibility period , so we’re unable to process a refund at this time.

You may click the link below to review our refund policy for more details.

15

After a long time since SUNO was unfortunately discontinued as a backend, I gave Tunee another chance.

Unfortunately, it failed completely this time as well. Tempo and specifications were ignored, as were layout and structure specifications. My specifications regarding the length of the track were also completely ignored, even though the model gave me feedback that it would generate a track adapted to my required track length.

Unfortunately, the sound quality was still rather poor. Everything sounded very muddy and undefined.

Conclusion: So far, it is absolutely unusable for my type of music. What a shame. :sleepy_face:

5 Likes
16

My Test :
Created an instrumental Intro in my DAW by using my Avenger2 VSTI-Synth. Uploaded this intro to tunee and asked to expand it to a full song and create lyrics.

This is what i’ve got :

17

I am patiently awaiting any model that is similar to the what we had before the Dec 31st upgrade. Tunee was the forerunner and reigning champion until then. I am still holding on though. You guys were just that good. We shall see.

3 Likes
18

Hi Tunee AI team,

I’ve been exploring the latest model update, and I wanted to share some thoughts. First of all, thank you for your work—the quality has definitely improved compared to how it was two weeks ago. It’s a step forward!

However, as a regular user, I still find it challenging to achieve the exact musical atmosphere I have in mind. To be honest, I really miss the specific magic and creative “logic” the model had before the January 1st update. That version felt more aligned with my artistic vision.

Nevertheless, I see the potential and I’m already trying to create new tracks with the current tools. I believe you’re on the right track! Wishing the whole Tunee AI team the best of luck with further improvements. Looking forward to seeing where you take us next!

Best regards, Limelim.

2 Likes
20

能不能设置一个翻译按钮，你这右上角的翻译对文本内容完全不管用啊。我想看帖子内容

21

你好，中国用户可以直接进入我们的微信社群，相应的内容也是有的哦~