claude dictation not working

The speech-to-text dictation feature in the Claude web interface and mobile applications offers a convenient hands-free way to draft prompts. However, when the microphone button fails to activate, is greyed out, or stops transcribing mid-sentence, it can break your productivity flow. This issue is typically caused by system permission blocks, browser audio conflicts, or network latency.

Grant browser and system microphone access

Both your operating system and web browser must explicitly permit the Claude web app or mobile application to access your microphone. If these permissions were previously declined, the dictation icon will remain inactive.

On a desktop browser, click the lock or settings icon in the address bar on claude.ai. Ensure that the "Microphone" toggle is enabled. If you are on an iOS or Android device, navigate to your system settings app, search for the "Claude" app, and verify that the microphone permission toggle is switched on. Without this system-level access, the app cannot receive audio input.

Select the correct default input device

If you have multiple audio devices connected to your computer, such as external monitors, webcams, or Bluetooth headsets, the browser may be trying to record from an inactive or disconnected hardware source.

Open your computer's system sound settings and verify which microphone is set as the default input device. Speak into the microphone to confirm that the system registers an active input level. Once verified, restart your browser to ensure it binds to the newly configured default audio source.

Clear app cache or reset web session

Corrupted cache files and temporary app bugs can prevent the speech recognition script from running smoothly, leading to the Claude app wont load loops or causing feature failures like dictation locks.

If you are using the mobile app, force close the application, clear its cache from your device's storage settings, and launch it again. If you are on desktop, perform a hard reload by pressing Ctrl + F5 (Windows) or Cmd + Shift + R (Mac). This ensures that the web app pulls down the latest speech recognition libraries from Anthropic's servers.

Switch to a supported browser engine

Claude's web-based dictation relies heavily on native browser Web Speech APIs. Not all browsers support these APIs equally or implement speech recognition in the same manner.

If dictation is failing on an alternative browser like Firefox or Brave, try switching to a Chromium-based browser (such as Google Chrome or Microsoft Edge) or Safari on macOS. These browsers have the most stable and native integration with cloud-based speech-to-text engines, which often resolves dictation bugs instantly.

Monitor network connection and latency

Unlike local dictation tools, web-based speech-to-text engines must stream your audio data to cloud transcription APIs to return real-time text. If your internet connection is unstable, the audio stream will drop, causing the transcription to fail or freeze.

Test your network speed to ensure your connection has low latency. If you suspect that a larger platform outage is causing the dictation feature to fail to process, check if the system is overall sluggish or explore the steps in our guide on Claude not responding to rule out broader server issues.