moeru-ai/airi · warning

Web Speech API recognition did not start immediately. This m

Error message

Web Speech API recognition did not start immediately. This might be due to:

What it means

This is a diagnostic `console.warn` (not an exception) emitted by `streamWebSpeechAPITranscription` when `startRecognition()` returned false, meaning the Web Speech API `recognition.start()` did not succeed immediately. It lists the three likely causes: missing microphone permission, an already-running recognition instance, or the browser requiring a user gesture. The stream is not rejected here; the code deliberately waits for permission or user action instead of retrying.

Source

Thrown at packages/provider-inference/src/providers/local/browser-web-speech-api/provider.ts:457

  }

  recognition.onsoundend = () => {
    console.info('Web Speech API sound ended')
  }

  recognition.onaudioend = () => {
    console.info('Web Speech API audio capture ended')
  }

  recognition.onnomatch = () => {
    console.info('Web Speech API: No speech match')
  }

  const started = startRecognition()
  if (!started) {
    // If immediate start failed, it might be a permission issue
    // Web Speech API will prompt for permission automatically, so we just log
    console.warn('Web Speech API recognition did not start immediately. This might be due to:')
    console.warn('1. Microphone permission not granted - browser should prompt automatically')
    console.warn('2. Recognition already running - this is normal if called multiple times')
    console.warn('3. Browser requires user gesture - ensure microphone was enabled by user action')

    // Don't retry immediately - wait for permission or user action
    // The recognition instance is already created, so it can be started later if needed
  }

  return {
    fullStream,
    text: deferredText.promise,
    textStream,
    recognition: recognitionInstance,
  }
}

View on GitHub (pinned to f679616c34)

Solutions

1. Trigger the transcription call from an explicit user gesture (button click) so the mic permission prompt can be shown and granted.
2. Check mic permission state with `navigator.permissions.query({ name: 'microphone' })` and guide the user to grant access in site settings.
3. Ensure only one recognition instance runs at a time; call `stop()` before starting a new session.
4. Check the DevTools console for the preceding 'recognition start failed' warning, which names the concrete cause.

Example fix

// before
const result = await streamWebSpeechAPITranscription(...)

// after
const perm = await navigator.permissions.query({ name: 'microphone' })
if (perm.state !== 'granted') {
  await requestMicPermissionViaUserGesture() // must run inside a click handler
}
const result = await streamWebSpeechAPITranscription(...)

const canUseSpeech = typeof window !== 'undefined' && ('SpeechRecognition' in window || 'webkitSpeechRecognition' in window)
const perm = await navigator.permissions.query({ name: 'microphone' })
if (perm.state === 'denied') throw new Error('Microphone permission denied')

function isSpeechApiSupported(): boolean { return typeof window !== 'undefined' && ('SpeechRecognition' in window || 'webkitSpeechRecognition' in window) }

try {
  const session = await streamWebSpeechAPITranscription(...)
} catch (error) {
  if (error instanceof Error && error.message.includes('permission')) {
    showMicPermissionHelp(); return
  }
  // non-fatal: start may simply be deferred — await session.text with timeout
  throw error
}

When it happens

Trigger: Calling `streamWebSpeechAPITranscription` (via `result`) in a browser where `recognition.start()` fails synchronously or asynchronously and the error is not classified as 'already started' or 'permission denied' by the handler, so `startRecognition` returns false.

Common situations: Calling transcription before any user interaction on a page that requires a gesture; the browser's mic permission prompt is pending or was dismissed; the recognizer was started twice without stopping the first instance; running in a browser/WebView without full Web Speech API support.

Related errors

• 2. Recognition already running - this is normal if called mu
• 1. Microphone permission not granted - browser should prompt
• Web Speech API error:
• Web Speech API: Microphone access issue. Please check microp
• 1. Microphone permission not granted - browser should prompt

AI-assisted analysis of moeru-ai/airi@f679616c34 (2026-09-08). Data as JSON: /api/errors/b3c772e2642f936b. Report an issue: GitHub.