Brizm
Independent Audio Analysis
Brizm builds audio analysis tools and APIs for developers, DJs, and producers. 100% independent — no venture capital, no data lock-in, no risk of the API disappearing to satisfy shareholders.
What Brizm Does
- 32 free audio tools that run on your device — BPM detection, key detection, stem separation, loudness metering, chord analysis, and more. Files stay local. Nothing is uploaded unless you choose cloud processing.
- A REST API that computes audio features from the actual waveform using current MIR architectures (2022–2024 era), not static database lookups from a decade ago. Float-precision tempo, benchmarked key detection, and continuously updated inference — every model ships improvements on a public changelog.
- The drop-in replacement for Spotify's deprecated Audio Features endpoint. Same field names, deterministic output, and full coverage for tracks that Spotify never catalogued.
Operating Principles
- Every score is computed from the audio waveform. No scraping. No caching someone else's numbers.
- Files processed locally or deleted immediately after cloud processing. Zero retention.
- Published benchmarks on standard academic datasets. Models continuously retrained and inference optimized as the field advances.
- Deterministic output — same audio in, same results out, every time.
9.3M+
Tracks Indexed
6
Production Models
32
Free Tools
94.8%
BPM Accuracy
Report bugs or request features on GitHub.