NEUTRINO

Overview

NEUTRINO turns written lyrics and melody into a synthesized singing voice, making it a strong fit for composers and producers who want vocals without recording a singer. Its no-cost music workflow is appealing, though application use needs separate permission and voice libraries come with their own terms.

It estimates timing, pitch, vocal quality, and breathiness from sheet music, then uses a vocoder to synthesize the part. Getting started means downloading the software and adding a voice library. The project says its models learn from recordings made with performers’ permission or audio released by public institutions, rather than from existing pretrained models.

Key features

Written-music vocal generation

NEUTRINO works from lyrics and melody, not vocal input. That makes it useful for building a sung line from a composition, but a poor match for converting an existing vocal performance. The project describes its models as reproducing qualities, singing styles, and quirks drawn from vocalists’ recordings.

Retakes, modes, and exports

Muon supports voice retakes and adjustable quality and processing speed, with advanced, standard, and elements modes. Advanced and standard render to 48 kHz, 16-bit mono WAV; elements produces 24 kHz, 16-bit mono WAV. The higher-resolution modes are the clearer fit when export specifications matter, while elements gives a lower-resolution option.

Voice-library limits

Users cannot currently build their own voice libraries with a released framework. That keeps NEUTRINO focused on the libraries provided for it and rules it out for teams that need to create a custom voice.

Hardware and distribution

The FAQ says about 4–6 GB of GPU VRAM is sufficient, though some processing may run out of memory below 4 GB; CPU processing modes are also described. As of May 2025, supported distribution services include JOYSOUND Uta-Suki Music Post, BIG UP!, TuneCore Japan, and Dwango’s music monetization service.

Pricing

NEUTRINO costs nothing for general music and audio production. That makes it a compelling option for composers and producers who want a singing synthesis workflow without a subscription. Embedding the software or using generated audio inside an application requires separate permission. NEUTRINO itself places no restriction on ordinary audio-file uses such as music production, but character-specific terms also apply and vary by character and asset type.

Platforms

NEUTRINO supports Linux, macOS, Windows, and web. Its download instructions describe a workflow that starts with downloading the software and adding a voice library. The project directs users to contact it if the FAQ does not resolve a support question.

Who it's for

NEUTRINO suits musicians who compose vocal parts from sheet music and want to render them as WAV without paying for ordinary audio production use. It is less suitable for anyone seeking vocal conversion, custom voice cloning, or a ready-made license for embedding generated audio in an application.

Pros and cons

• Pro: Free use for general music and audio production makes it accessible for composing and producing vocal parts.
• Pro: Retakes, quality and speed controls, and three processing modes give users ways to adjust generation.
• Pro: Advanced and standard output at 48 kHz, 16-bit mono WAV meets a clear, stated export specification.
• Con: Application integration and in-app use of generated audio require separate permission, adding a hurdle for software makers.
• Con: Character-specific terms vary by character and asset type, so ordinary audio-use permission does not settle every use case.
• Con: Users cannot create their own voice libraries with a released framework, limiting custom-voice work.
• Con: Some processing may run out of memory below 4 GB of GPU VRAM.

Alternatives

For a broader set of platforms and a free tier capped at five conversion minutes per month with one Starter AI voice, consider Revocalize AI. Developers and researchers who want code and model weights under an Apache 2.0 license can choose SoulX-Singer. VoiSona Song is another freemium singing option for iOS, macOS, and Windows.

Kits AI has a free tier with 15 conversion minutes per month and voice-generation tools. Audimee offers 11 royalty-free voices and 15 free conversion minutes, with MP3 downloads and personal-use-only terms. LyricToMelody AI starts with 20 free credits for lyric-first drafts and exports to MP3, WAV, or PCM, retaining projects for seven days. Tunela bundles limited free conversions with stem separation, text-to-speech, and video swaps. SightSinger is a web-based freemium alternative with a free trial.

Browse the AI Singing Voice Generators category for more options.

Verdict

Choose NEUTRINO if you write vocal parts as music and want a free way to turn them into synthesized singing, with WAV output and adjustable generation modes. Look elsewhere if you need vocal input, a custom voice library, or application use without securing separate permission.