Neiro

Overview

Neiro is a locally processed audio suite that brings together source separation, restoration, transcription, editing, and music practice. It is best suited to musicians and audio editors who want those jobs in one workflow and are comfortable managing optional models and software dependencies. Its breadth and local processing are compelling; the trade-off is that neural models bring separate licensing questions, and some media formats need ffmpeg.

Key features

Separation and repair

Neiro can isolate vocals, instrumentals, harmonic and percussive parts, four- or six-stem mixes, and drum kits. A null-test residual accompanies each separation result, giving users a way to inspect what remains rather than treating the split as a black box. The repair toolkit covers declipping, hum and noise removal, click and crackle removal, dereverberation, bandwidth extension, and reference mastering; batch processing is supported. That range makes it useful for both mix analysis and cleanup, though the neural options are not a turnkey part of the desktop download.

Transcription, editing, and practice

Audio-to-MIDI transcription is complemented by MusicXML, ASCII tablature, and LRC lyric exports. Studio's non-destructive waveform editing keeps changes reversible, while Learn provides loop regions, count-in, metronome, step mode, WebMIDI, and DAW wait mode. This combination suits musicians moving between listening, transcription, editing, and practice; it is less compelling for someone who only needs a single-purpose restoration tool.

Local workflow and extensions

Neiro processes audio on the user's machine. Its desktop app is built with Tauri, and a browser interface can be launched with neiro ui; that interface binds to 127.0.0.1, while the desktop shell restricts connections to the local engine. The security policy says there is no outbound network activity by default beyond user-initiated model downloads and updates. Model downloads are checked against manifest SHA-256 values, but third-party weights can still be dangerous. VST2 and CLAP injectors, plus a VST2 pass-through effect, let users capture audio from a DAW into Neiro. A documented Python adapter plugin MVP runs granted adapters in the Neiro process without a sandbox, so extensions deserve caution.

Pricing

Neiro is free, with a free plan and no paid plan described. The core DSP floor works without model downloads, but neural backends such as Demucs, Basic Pitch, and AudioSR are optional and download weights on first use; those models carry their own licenses, including some non-commercial or research-only terms. The engine, desktop shell, and frontend are MIT licensed, which does not override the terms attached to individual models. Users who need those neural capabilities should check each model's license for their intended use.

Platforms

Neiro supports Linux, macOS, Windows, web, and self-hosted use. Desktop releases provide Windows MSI/EXE, macOS DMG, and Linux AppImage/DEB installers. The Python package requires Python 3.10–3.12. WAV and FLAC work without ffmpeg, but compressed or video inputs require ffmpeg on PATH, an extra setup step for users handling those formats.

Who it's for

Neiro makes the most sense for musicians, producers, and audio editors who want local processing across separation, repair, transcription, and practice, especially those already comfortable with a desktop or self-hosted workflow. Its free price and model-free DSP floor lower the barrier to trying core functions. It is a weaker fit for teams seeking a supported commercial model catalog without license-by-license checks, or for users who need compressed and video inputs to work without installing ffmpeg. Full MUSDB18-HQ and MAESTRO evaluation numbers require users to provide those datasets, so benchmarking those tasks also takes extra effort.

Pros and cons

• Pros: Broad audio workflow in one locally processed suite, from source separation and repair through transcription, editing, and practice.
• Pros: Core DSP functions work without model downloads, and the app and interface are free.
• Pros: Separation results include null-test residuals, and DAW capture is supported through VST2 and CLAP injectors.
• Cons: Neural weights are separate downloads with model-specific licenses, some restricted to non-commercial or research use.
• Cons: Compressed and video inputs require ffmpeg on PATH; WAV and FLAC avoid that dependency.
• Cons: Granted Python adapters run without a sandbox, and third-party model weights carry a security warning.

Alternatives

Audio Restoration Software is the category guide for comparing restoration tools. Consider Wave Arts Master Restoration Suite 6 if you want five restoration plug-ins for a host that supports audio plug-ins: it costs 99.00 USD per once and has a free trial, but is paid software for macOS and Windows. CEDAR Cambridge is a paid modular hardware and software system whose options and configuration vary; contact CEDAR Audio for information.

SpectraLayers Pro is another paid macOS and Windows option with a free trial, focused on spectral audio editing, repair, restoration, and AI-assisted processing. Choose LANDR ReHance if its ReHance tool within LANDR Studio suits your workflow: the subscription is 11.99 USD per month on annual plans, and it is also available on the web. Cathar is a free alternative for Linux and macOS.

Diamond Cut Audio Restoration Tools 11.09 is a Windows-only paid option at 59.00 USD per once, with a free trial and one year of free support. Acoustica is a paid macOS and Windows alternative with Standard and Premium plans; Standard is described for two-channel stereo, while Premium supports up to 7.1.2. iZotope RX is a paid macOS and Windows suite with Elements and Standard tiers, for readers comparing a dedicated repair-plugin lineup.

Verdict

Choose Neiro if you want a free, local audio workspace that spans separation, restoration, transcription, editing, and practice without requiring neural models for its core DSP functions. Its central advantage is that breadth in one locally processed tool; look elsewhere if you need a simpler dependency setup or cannot work within the separate licenses and security considerations attached to optional models.