Skip to main content
Use this page to troubleshoot common issues in BeyondWords. If you don’t find what you need here or elsewhere in these docs, contact our support team.

​
Voice issues

If the voice is pronouncing a word or phrase incorrectly, add a custom pronunciation to fix it.If the issue persists, check:
  • The pronunciation is applied at the correct level (content, project, or organization)
  • The content has been regenerated after adding the rule
If the voice sounds unnatural or doesn’t match your expectations, try:
  • Switching to a different voice model
  • Adjusting voice settings such as stability, style, or similarity
  • Using a different voice (premade, generated, or cloned)
For best results, choose a voice that matches your content type and language.

​
Content extraction issues

If unwanted content was extracted from your page, there are three possible solutions: 
  1. Add an Exclude filter in your content extraction settings and regenerate the content. This is the best option for HTML elements that should never be included (e.g., captions).
  2. Edit and regenerate the content via the Editor. This is the best option for one-off issues. To prevent future edits to the live article from overwriting your changes, disable API updates in the Info tab.
  3. Fix any Include filters that are unintentionally extracting the unwanted content, then regenerate your content.
If content that should be extracted is not being extracted, there are two possible solutions: 
  1. Add an Include filter in your content extraction settings and regenerate the content. This is the best option for HTML elements that should never be included (e.g., captions).
  2. Fix any Exclude filters that are unintentionally excluding the required content, then regenerate your content.
If a content filter is not working, it may be due to one of the following reasons:
  • The content uses the manual_segment processing type. This occurs when the content was created or last edited in the Editor. In this mode, content extraction filters are not applied. There are three possible ways to resolve this issue:
    • Make the required changes directly in the Editor and regenerate the content
    • Send a new API update with the processing type set to auto_segment (please confirm with your development team before attempting)
    • Regenerate the content via the API and avoid editing it in the Editor
  • The filter is configured incorrectly or conflicts with another. Review your filters and page markup to ensure the correct HTML elements are being targeted.
  • The content has not been regenerated since the filter was added or updated. (Note that dashboard regeneration will not apply content extraction filters if the content was created or last edited in the Editor. In this case, the content must be updated via the API.)
If your article is password-protected or behind a paywall, BeyondWords may extract the login or password prompt page instead of the article content.To grant BeyondWords access to protected content, you can either: 
  • Set request headers: Go to Settings → Extraction and add the relevant authentication headers
  • Enable static IP: Go to Settings → Extraction, switch the Static IP toggle on, and allowlist the displayed IP addresses in your firewall, CDN, or server configuration

​
Editor issues

If your content is created via an integration, updates to the live article will automatically sync to the content item. This means any manual edits you make in the Editor may be overwritten.You can prevent this by disabling API updates for the item:
  1. Open the content in the Editor
  2. Open the Info tab in the left-hand sidebar
  3. Scroll down to disable API updates
  4. Click Regenerate to save changes
Once API updates are disabled, changes to the original article will no longer be reflected in generated audio and video. Also, new content filters will not be applied.

​
Audio issues

If audio generation has stopped, check the following:
  • You have available credits remaining
  • There are no errors in your content (e.g., unsupported formatting)
You can also try regenerating the content.If the issue continues, contact support with the content ID and any error messages.
If audio is shorter than expected or missing sections:
  • Check whether the source content was fully extracted
  • Look for skipped or excluded segments in the Editor
  • Ensure there are no formatting issues affecting segmentation
Updating and regenerating the content will usually resolve this.
If audio sounds too fast or too slow: 
  1. Adjust the speaking rate in your voice settings
  2. Regenerate the content after making changes
Slower speaking rates will result in longer audio, while faster rates will shorten the duration.

​
Player issues

If the player isn’t appearing on your site, confirm the following
  • The content has fully generated
  • Automatic publishing is enabled in project settings
  • The player is installed correctly
  • The content’s visibility is set to public
If the player loads but doesn’t play:
  • Check your browser console for errors
  • Confirm the audio has finished generating
  • Ensure there are no network or autoplay restrictions
Try opening the content via its shareable URL to confirm playback works outside your site.
If content doesn’t continue playing automatically:
  • Check your continuous playback settings in the dashboard
  • Ensure there is additional content available to queue
  • Confirm your player configuration supports continuous playback
Note that behavior may vary depending on your player setup and settings.

​
Other issues

If the content item’s title is incorrect, go to the Info tab in the Editor and manually update the Meta title field.(When the first segment is created, BeyondWords uses this text as the meta title. If this first segment is later changed, the meta title does not automatically synchronize with the updated text. This can cause the spoken title and meta title to differ.)