Skip to main content

Speech-to-Text

PromptCue’s Speech-to-Text feature allows you to speak naturally into your microphone, converting your voice into text. Powered by the Web Speech API, this feature is designed to make your interactions with AI models even more fluid—whether you prefer typing or speaking.

How It Works?​

  1. Voice Input:
    Simply click the microphone icon in the chatbox and begin speaking. Your voice is captured through your device’s microphone.

  2. Real-Time Transcription:
    As you speak, the Web Speech API converts your speech into text in real time. When you stop recording, the transcribed text appears in the chatbox, ready for you to review and send.

  3. Seamless Integration:
    The transcribed text is seamlessly integrated into your conversation, allowing you to edit if needed before sending it off to your selected AI model.

Supported Browsers and Compatibility​

The Web Speech API is supported in several modern browsers—most notably in Chrome and Safari. However, there are known compatibility differences:

  • Chrome:
    Generally offer robust support for the API, though some users have reported intermittent false network errors and other quirks.
  • Safari:
    Offers a more stable experience with speech recognition, often delivering more reliable results.
  • Firefox & Others:
    The Web Speech API is either not supported or only partially supported, so for the best experience, we recommend using Chrome or Safari.

Limitations & Common Errors​

While the Speech-to-Text feature enhances user interaction, the underlying Web Speech API comes with some limitations. Here are a few common error types and known issues:

  • not-allowed:
    Indicates that the user has not granted microphone permissions.
    Solution: Ensure that your browser has permission to access the microphone.

  • language-not-supported:
    Triggered when the selected language is not supported by the API.
    Solution: Verify the language settings in your browser.

  • no-speech:
    Occurs when no speech is detected during the recording.
    Solution: Ensure you speak clearly and that your microphone is functioning properly.

  • network:
    A network error that can sometimes occur even when you are online, particularly in Chrome/Brave.
    Note: This error is a known quirk of the API. Our system attempts to handle it gracefully, but if it persists, try restarting the recording.

  • audio-capture:
    Means no microphone was detected on your device.
    Solution: Check that your device has a functioning microphone and that it’s connected.

  • aborted:
    Indicates that the recognition process was aborted—often because the user manually stopped the recording.
    Solution: This error is usually expected during normal operation when you click the stop button.

Additional Limitations​

  • Session Duration:
    The Web Speech API is designed for short bursts of speech. Long continuous speech may be cut off or result in errors.
  • Inconsistent Accuracy:
    Accuracy can vary based on background noise, microphone quality, and accent.
  • Browser Variability:
    Not all browsers implement the API equally; some may have additional quirks or less reliable performance.
Best Practices
  • Ensure Permissions:
    Before using the feature, grant your browser permission to access the microphone.
  • Speak Clearly:
    To improve transcription accuracy, try to speak clearly and minimize background noise.
  • Use Short Bursts:
    Keep your voice input concise to avoid issues with session timeouts.
  • Review Transcriptions:
    Always review the transcribed text before sending it to the AI model to ensure accuracy.
  • Fallback Option:
    If you encounter persistent errors, switch to typing your prompt as an alternative.

Next Steps​

After understanding the Speech-to-Text feature, you can further explore:

  • Chatbox – Learn more about how to interact with PromptCue via text and voice.
  • API Key Setup – Ensure your API keys are configured to enable seamless AI interactions.
  • Supported AI Models – Review the range of AI models you can access with PromptCue.

Experience a new level of interaction—speak your ideas, and let PromptCue handle the rest!