IBM Watson Speech JavaScript SDK Examples

The watson-speech library allows you to easily add voice recognition and synthesis to any web app with minimal code.

Complete source code for these examples is available on GitHub.

Speech to Text

Microphone Input

• Transcribe from Microphone
• Transcribe from Microphone, with pre-loaded token (compatible with iOS)
• Transcribe from Microphone, automatically stop at first pause
• Transcribe from Microphone, with chosen model
• Transcribe from Microphone, with Alternatives
• Transcribe from Microphone, with Word Confidence
• Transcribe from Microphone, send text to console
• Transcribe from Microphone, send JSON to console (includes text and metadata; v0.22+ format)
• Browserify example
• Webpack example
• Deprecated: Transcribe from Microphone, send JSON to console with results extracted (pre-v0.22 format)

File Input

(user-provided or downloaded - several sample files available in static/ folder)

• Transcribe and Play a file via URL, with pre-loaded token (compatible with iOS)
• Transcribe and Play a user-provided file
• Transcribe from URL with {speaker_labels: true}, output to console
• Transcribe from URL with {resultsBySpeaker: true}, output to console
• Transcribe from URL with {resultsBySpeaker: true}, output HTML
• Compare realtime and TimingStream options
• Transcribe from URL, return a Promise
• Transcribe from file, pre-loaded over AJAX
• Deprecated: Transcribe from HTML5 <audio> or <video> element

Text to Speech

• Synthesize text
• Synthesize text, with pre-loaded wav audio (works in Edge)
• Synthesize text w/ custom voice

See Also

• JSDoc
• npm
• GitHub
• Official Speech to Text demo
• Official Text to Speech demo
• Watson