StemForge

Overview

Separate, MIDI, Synth, Compose, Mix and Export are connected stages: outputs can flow between them, and the Mix tab can combine separated audio, MIDI-rendered tracks, synth outputs and composed songs. That makes StemForge more useful for projects that move across several audio tasks than for someone who only needs a quick vocal split or remix.

The project says audio stays on the user's computer and that an internet connection is unnecessary after initial model downloads. That is a strong fit for people who want local processing, but it does not eliminate the initial download burden or the need to run the software on suitable hardware.

Key features

Separation and MIDI

Demucs and BS-Roformer separate vocals, drums, bass, other, guitar and piano. StemForge recommends engines and models based on spectral analysis, giving users guidance when choosing a separation path. For MIDI, BasicPitch extracts notes from instrument stems, while faster-whisper and pitch tracking handle vocal melodies. MIDI Out can also send material to hardware synths through Web MIDI, extending the workflow beyond files on the computer.

Generation, composition and export

Synth uses Stable Audio Open for text-conditioned audio generation up to 600 seconds, with optional audio or MIDI conditioning. Compose combines AceStep song generation with RVC voice conversion, and includes 14 voices plus a HuggingFace model browser. These options make the workstation broader than a dedicated stem splitter, though model licensing and downloads deserve attention before relying on a particular generation workflow.

Mix brings together stems, MIDI-rendered tracks, synth outputs and composed songs. Export supports WAV, FLAC, MP3 and OGG, either as individual files or a ZIP archive. That range covers common delivery formats without requiring a separate export utility for each output.

Pricing

StemForge is free: the StemForge plan costs 0.00 USD per free, with no subscriptions or per-track fees. It is open source under Apache 2.0 and runs on the user's hardware, so there is no paid tier or seat-based subscription to weigh against the free option.

The trade-off is that free software does not mean cost-free setup. The documented requirements include Python 3.12, FFmpeg and FluidSynth with a GM soundfont. NVIDIA acceleration requires driver 580 or newer, and CPU-only operation is significantly slower. Models download on first use; AceStep models require about 20 GB, while Stable Audio Open requires HuggingFace license acceptance and authentication. Upstream model terms are separate from StemForge's Apache 2.0 license: the project notes Stability AI's Community License for Stable Audio Open and no specified license for one six-stem BS-Roformer model.

Platforms

StemForge supports Linux, macOS and Windows, with a local browser interface and a self-hosted option. MIDI Out has an additional browser constraint: it requires Chrome or Edge in a secure context, such as HTTPS or localhost. Users planning to connect hardware synths should account for that requirement.

Who it's for

StemForge suits technically confident musicians, producers and audio tinkerers who want several connected audio workflows, local processing and the flexibility to run models on their own machine. It is a poor fit for people who need a lightweight setup, depend on fast CPU-only processing, or want to avoid large downloads and model-specific licensing terms.

Pros and cons

Pros

• Connected separation, MIDI, generation, composition, mixing and export workflows reduce the need to move every stage between separate tools.
• Local processing after downloads and no subscription or per-track fees suit users who prefer to keep audio on their hardware.
• Six stem categories, MIDI extraction, Web MIDI output and four export formats support more than a basic vocal-removal task.

Cons

• Python, FFmpeg and FluidSynth setup, plus much slower CPU-only operation, make this a demanding choice for casual users or modest hardware.
• AceStep's roughly 20 GB model requirement and first-use downloads can consume substantial storage and time.
• Apache 2.0 covers StemForge, not every upstream model; separate terms and an unspecified license for one BS-Roformer model complicate model selection.
• MIDI Out is limited to Chrome or Edge in a secure browser context, so it will not work from every browser setup.

Alternatives

For a browser- and app-accessible freemium option, Fadr offers a free plan with vocals, melodies, drums and bass, MP3 downloads, MIDI detection and Remix Maker; choose it if those focused remix tools are a better fit than StemForge's local setup. AceStep AI Music Remixer has a free tier with 10 credits a day and generations up to 120 seconds for personal use, so it suits users who want a capped, shorter-generation option rather than StemForge's local workstation.

AetherWave Music Studio, AI Remix Music Generator, MeloLab AI Remix, Mubert AI Music Remixer, MusicWave.ai and RiffGen Remixer AI are other alternatives to consider.

For a broader comparison of category options, see AI Music Remixing Tools.

Verdict

Choose StemForge if you want a free, locally run workstation that connects stem separation with MIDI, generation, composition and export. The integrated workflow and lack of recurring fees are its clearest strengths. Look elsewhere if you need a quick-start service, have limited hardware or storage, or cannot work within separate upstream model licenses.