Basic Pitch

External

Basic Pitch by Spotify is a free, open-source audio-to-MIDI converter that excels at transcribing single-instrument recordings, including vocals, with precise pitch bend detection for elements like vibrato and slides. It processes audio entirely locally in your browser for complete privacy, using a lightweight model that's under 20MB and runs up to 10x faster than real-time. Musicians, producers, educators, and developers love it for quickly turning hums, riffs, or performances into editable MIDI files importable into any DAW, without needing extra hardware.

CategoryMusic & Audio
Basic Pitch

Description

Basic Pitch by Spotify is a free, open-source audio-to-MIDI converter that excels at transcribing single-instrument recordings, including vocals, with precise pitch bend detection for elements like vibrato and slides. It processes audio entirely locally in your browser for complete privacy, using a lightweight model that's under 20MB and runs up to 10x faster than real-time. Musicians, producers, educators, and developers love it for quickly turning hums, riffs, or performances into editable MIDI files importable into any DAW, without needing extra hardware.

Key capabilities

  • Audio-to-MIDI conversion with pitch bend detection
  • Polyphonic transcription for single instruments and voice
  • Local browser-based processing with no uploads
  • Lightweight model (under 20MB, 17k parameters, 10x real-time speed)
  • Supports WAV, MP3, OGG, FLAC, M4A inputs
  • Open-source Python/NPM library (Apache-2.0)

Core use cases

  1. 1.Quickly transcribe single-instrument ideas or hums to MIDI
  2. 2.Capture and edit real performances in a DAW
  3. 3.Teach audio transcription to beginners without MIDI keyboards
  4. 4.Integrate automatic music transcription into applications

Is Basic Pitch Right for You?

Best for

  • Musicians/producers capturing single-instrument riffs or vocals
  • Beginners/educators transcribing performances easily
  • Developers building lightweight AMT features

Not ideal for

  • Users separating multiple instruments in mixes
  • Professionals needing flawless, edit-free transcriptions

Standout features

  • Drag-and-drop audio upload or built-in recording
  • Instant MIDI preview and download
  • High accuracy on pitch bends, vibrato, and slides
  • Efficient on-device inference
  • Multiple audio format support
  • Adjustable thresholds for note segmentation and confidence

User Feedback Highlights

Most Praised

  • High accuracy for single-instrument polyphony and pitch bends
  • User-friendly demo with instant results
  • Outperforms some larger models in note accuracy
  • Lightweight and fast for everyday use
  • Popular GitHub repo with 4.6k stars and active maintenance

Common Complaints

  • Struggles with dense multi-instrument mixes
  • May produce repeated or imprecise note lengths
  • Lower accuracy on highly complex or choppy solos
  • Requires post-edits in DAW for professional polish