iOS Safari: SpeechRecognition + audio playback loop breaks after first cycle

Issue: iOS Safari Web Speech API + audio playback loop breaks after first cycle

I’m building a browser-based voice interaction loop:

TTS playback → start SpeechRecognition → process result → play TTS again → repeat

Tech:

- webkitSpeechRecognition (Web Speech API)

- <audio>.play() for TTS (ElevenLabs)

- Plain JS (no framework)

Problem (iPhone / iOS Safari & Chrome):

- First cycle works correctly

- On subsequent cycles:

- audio playback fails or is silent

- or SpeechRecognition gets stuck / doesn’t restart

- sometimes a “click” occurs with no audio output

Desktop browsers work fine.

Suspected cause:

iOS WebKit audio session conflict when switching repeatedly between:

- microphone input (SpeechRecognition)

- audio output (HTMLAudioElement)

Question:

Is this a known limitation of iOS WebKit?

Is there any reliable way to alternate mic + playback in-browser, or is the only stable approach:

- continuous recording (getUserMedia + MediaRecorder), or

- server-side transcription (e.g., Whisper), avoiding SpeechRecognition entirely?

0 comments

skool.com/quantum-quill-lyceum-1116

Jake Van Clief, giving you the Cliff notes on the new AI age.

Leaderboard (30-day)

+731

+548

+465

+406

+386