MAGDA is a free, open-source digital audio workstation focused on AI-assisted music production, automation, transformation, and fast musical iteration. It addresses the needs of musicians and producers who want to combine traditional DAW workflows with AI-driven tools for mixing, sound design, and creative generation.
The application offers a clip-based session view for jamming, a timeline arrangement for full productions, and a dedicated mixer. Mixing and analysis capabilities include one-click mix analysis with per-track loudness, dynamics, and masking detection plus a mixing agent that interprets the results. Additional tools comprise an oscilloscope, spectrum analyzer, levels meter, and automatic gain staging. Its instruments and effects feature a FAUST-powered bank including filter, compressor, reverb, saturator, clipper, and modulation modules, the 4OSC subtractive synthesizer, a drum grid with per-pad plugin chains, and an experimental FAUST device for writing custom DSP.
AI agents can access session context, generate commands through a built-in DSL, and design sounds on devices such as the 4OSC. These agents support cloud providers with user-supplied API credentials or fully local operation via llama.cpp or a local server. Automation options allow drawing curves, recording parameter moves in write, touch, or latch modes, or using AI to generate them. Hardware control is handled through controller profiles, MIDI Learn, and Lua scripting. The media library is indexed and searchable by family, key, tempo, tags, or plain-language descriptions that enable AI similarity ranking, with direct drag-and-drop import.
Further functions include parameter customisation that lets users hide unused controls, assign units and ranges, or apply AI classification, with a learn mode that identifies controls by clicking in the plugin window. Creative MIDI devices consist of a chord engine for real-time detection and suggestions, an arpeggiator, and a self-driving step sequencer with step recording and AI prompt-to-pattern generation. Time-bend and groove features redistribute notes along curves or apply non-destructive feel templates. Modulation operates at sample rate using LFOs and Bezier curve shapers at device, rack, or track level, while macros can control other modulators. Racks support multi-chain parallel and serial signal paths, and presets can be saved for devices or entire track chains; the software automatically scans existing .vstpreset and .aupreset libraries. Tracks handle hybrid audio and MIDI, with bounce, freeze, flatten, async BPM detection, and QWERTY MIDI input available.
MAGDA is built on the Tracktion Engine, uses JUCE for its interface, and employs FAUST for DSP. It runs on macOS with native Apple Silicon support, Windows, and Linux; Intel Macs receive limited support without AI sample search and are expected to lose compatibility after the current version series.
In the Audio editing & DAW space, MAGDA takes a focused approach. It enables musicians and producers to create, arrange, and mix music efficiently with AI-assisted tools and automation. It is built as an open-source project for musicians and music producers. MAGDA is open source under the GPL-3.0 license. It runs on the web, macOS, Windows, and Linux.
Behind MAGDA is Conceptual Machines, and it first shipped in 2025. The project is developed in the open on GitHub with 165 stars and 137 commits in the last 90 days. Key capabilities include session view, arrangement view, and mixer.
Summary written by a language model from the project’s public pages.
What PulseGate has recorded for this listing
Same category — not a similarity match