Soundverse

Freemium

Soundverse is an AI music generation studio. You type what you want and get stems, full tracks, or vocals back.

It packs a lot in: text-to-music, stem separation, voice cloning, lyrics generation, and a multi-track editor. You can hum a melody and have it turn into an arrangement. The AI agent asks you questions and builds the track through conversation.

The tool sits in music generation. It is built for musicians, content creators, and hobbyists who want AI-made audio without learning a full DAW.

Open it if you want to sketch ideas and slice up AI output. Skip it if you need reliable tools that work the first time.

It helps most when you have patience for broken features and slow results.

• Podcasters needing intro music mixed with voice tracks: You can generate a theme and drop it under your voice in the same tool. No round-trip to another editor.
• Short-form video creators: You need quick background loops for TikTok or Reels. The mood-matching works when you plan for a few tries.
• Producers using AI as a sketchpad: You start demos here and finish them in Logic or Ableton. The stem export lets you pull pieces out.
• Users who prefer detailed prompt control: You write long prompts with mood, BPM, and instruments. The output follows closer when you put the work in upfront.
• Creators comfortable editing and slicing AI output: You treat the first generation as raw material. You cut, loop, and layer it yourself.
• Those working in pure single genres: Pop, rock, lo-fi, and metal come out clean. Fusion styles fall apart.
• Users willing to pay without testing: The free tier hits limits fast. You buy tokens or subscribe to get real work done.
• Those planning monthly subscriptions: Lifetime deal holders report being locked out of updates. Monthly users get better support.

These parts work well enough that you keep coming back.

• Stem separation: You can pull vocals, drums, bass, and other instruments into separate tracks. It is useful for remixing even if the separation gets muddy.
• Inpainting: You regenerate just a section of a track instead of starting over. It saves time when most of the song is right.
• Conversational agent workflow: You describe what you want in plain chat. The AI asks follow-ups and remembers context. It feels less like coding and more like talking to a collaborator.
• AI singing generator: You feed it lyrics and get vocal tracks back. It works best for short hooks and jingles under 16 seconds.
• Simple and Custom modes: You can type a quick sentence or get deep with lyrics, style, and reference uploads. You pick the depth you need.

The path has more steps than it should.

1. Open the web studio and describe your track in the chat with Agent One.
2. Upload a reference or hum a melody if you want the AI to follow your idea.
3. Choose Simple mode for a fast result or Custom mode for lyrics, genre, and style control.
4. Generate multiple options and pick the closest match.
5. Export stems to finish in your DAW, or edit in the built-in multi-track timeline.

The problems run deep enough that you might lose money and time.

• Broken features: MIDI import gets treated as audio. Melody preservation fails. Humming-to-song changes your tune beyond recognition.
• Free tier is a trap: You hit rate limits after one to three songs. Documentation is buried and useless. Paid tokens sometimes do not unlock anything.
• Support is missing: Emails go unanswered for months. Refund requests are ignored even inside the stated window.
• Paid users get charged for failures: You pay even when generations fail or produce nothing. The token system is confusing and hard to forecast.
• Interface freezes: The tool crashes mid-session. You lose your place and burn tokens.
• Lifetime deal holders lose access: Users who bought Creator or Pro tiers report being downgraded to free. Promised updates never arrive.