Web Audio APIwithout browser

For simple graphs, yes: every measured scenario renders hundreds of times faster than realtime. The native Rust binding stays 1.5–3× quicker; AudioWorklet execution is currently synchronous, so production low-latency work remains the boundary.

@audio/speaker provides the default output device. Microphone examples use the optional @audio/mic adapter, then feed PCM into a MediaStreamTrack. You can also set sinkId to any Node writable stream.

decodeAudioData() uses @audio/decode. Its codec and container coverage is broader and more predictable than the native decodeAudioData() set in a typical browser, which varies by browser and operating system. Codecs are detected and loaded on demand for MP3, WAV, Ogg Vorbis, Opus, FLAC, AAC, ALAC, AIFF, CAF, QOA, WebM, WMA, AC-3, DTS, tracker modules, DSD, and more. Supported tracks can also be extracted from MP4, MOV, MKV, WebM, and AVI without FFmpeg or native bindings. Codec packages carry their own licences.

The core is 208 KB unpacked. Codecs load on demand through @audio/decode, TypeScript types ship in the package, and the project has been maintained since 2013.

Yes. Install the globals before dynamically importing Tone.js so its browser checks see the Web Audio constructors.

import 'web-audio-api/polyfill'
const Tone = await import('tone')

const ctx = new AudioContext()
await ctx.resume()
Tone.setContext(ctx)
new Tone.Synth().toDestination()
  .triggerAttackRelease('C4', '8n')

Yes. Render in memory, then assert samples, peaks, RMS, spectra, duration, or channel count. No audio device is opened.

import assert from 'node:assert/strict'
import { OfflineAudioContext } from 'web-audio-api'

const ctx = new OfflineAudioContext(1, 44100, 44100)
const osc = ctx.createOscillator()
osc.connect(ctx.destination)
osc.start()

const audio = await ctx.startRendering()
assert(audio.getChannelData(0).some(Boolean))

Yes. OfflineAudioContext renders without opening @audio/speaker. A realtime context can instead send PCM to a writable stream through sinkId.

import { OfflineAudioContext } from 'web-audio-api'

const ctx = new OfflineAudioContext(2, 44100, 44100)
// build the same graph here
const audio = await ctx.startRendering()

Yes. Modules can be registered by URL, data URL, Blob URL, or an inline callback in Node. Worklets currently execute synchronously rather than on an isolated realtime thread.

import { AudioContext, AudioWorkletNode } from 'web-audio-api'

const ctx = new AudioContext()
await ctx.resume()
await ctx.audioWorklet.addModule(scope => {
  class PassThrough extends scope.AudioWorkletProcessor {
    process(inputs, outputs) {
      outputs[0][0].set(inputs[0][0] || [])
      return true
    }
  }
  scope.registerProcessor('pass-through', PassThrough)
})

const node = new AudioWorkletNode(ctx, 'pass-through')

The I/O boundary. Node has no built-in speakers, microphones, or media elements: realtime output goes through @audio/speaker or any writable stream via sinkId, microphones come from @audio/mic, and files decode through @audio/decode. AudioWorklets run synchronously instead of on an isolated realtime thread. Everything else follows the same specification: the graph API, the 128-frame render quantum, automation, and node behavior.