TTS responses with word alignment

[rose-m]

Hey there,

I wanted to reach out and ask if there's a way to get word alignment information for the generated audio from the TTS API? So when providing "This is a test" we'd like to know at which point in the audio "This" is basically spoken (start / end) - ideally this would be on individual character level.

I didn't find that info anywhere in the docs for now... Mabye you can help me out there?

[team-deepgram]

Thanks for asking your question about Deepgram! If you didn't already include it in your post, please be sure to add as much detail as possible so we can assist you efficiently, such as:

• The request_id if you have a question about your requests or transcription responses.
• The features you used or the full api.deepgram.com URL you sent your request to, including parameters.
• Any code snippets you can share.

[jkroll-deepgram]

Hi @rose-m, we don't currently offer word alignment info on our TTS output audio.

If word timings are essential, you can transcribe the audio with our pre-recorded STT, which does provide word-level timestamps.

However, that would incur an additional transcription cost, which may not be acceptable for your use case.

[rose-m]

Thank you for your response - is that on the timeline / planned as a feature?

[jkroll-deepgram]

Hi @rose-m, we have heard that request from customers, and while it's not scoped out on our roadmap, it is a feature we are considering adding in the future. I've logged your interest on that feature, so we can reach out when we have additional information on our plans.

[PeachyCard]

Right now we are using 11Labs because they support timestamps on TTS. If you had this functionality, then we would leave 11Labs as your latency is much better