I’m trying to build a genuinely great “listen to this post” experience for my Ghost publication and would love to hear what people are actually using in production today.
I’ve tested ElevenLabs Audio Native and I’m also looking at BeyondWords. I’m not tied to either one. I’m trying to understand what the best current architecture is for Ghost.
What I’m looking for:
- Very high-quality, natural TTS
- Narration of the actual post body only
- Automatic generation when a post is published or updated
- A clean embedded audio player
- Support for cloning the author’s voice
- Minimal manual work per article
- Good performance on mobile
I’m open to BeyondWords, ElevenLabs, a custom Ghost integration using webhooks/API, or a newer TTS service/model if there’s a better option.
For anyone who has built this successfully on Ghost: what are you using, how is the workflow set up, and would you choose the same stack again today?
I’m especially interested in what you consider best-in-class, not just what technically works.
You can plug Ghost into BeyondWords or ElevenLabs or pretty much any other generator, either automatically with a webhook, or with a little cloud function sitting somewhere.
The big thing to watch out for is not to blow out your budget – be careful not to regenerate when you edit the post (unless you really really need to), and think carefully before letting any service do your whole back catalog.
2 Likes
Thanks Cathy, that’s useful, especially the regeneration warning.
One clarification: the ElevenLabs link is for ElevenAgents, the conversational widget, not Audio Native. Different product. Worth flagging since it’s the page that ranks for “ElevenLabs Ghost.”
More context on my situation. I have an ElevenLabs account with a Professional Voice Clone I’ve been training for about six months, and inference cost is not a constraint for me. So the clone is the asset I’m protecting, not the per-post spend.
That reframes it. Two questions:
Can BeyondWords narrate using my own ElevenLabs Professional Voice Clone via my API key, or am I limited to their voice catalog? If it’s my clone, the official integration is the obvious answer.
If not, has anyone here built the direct path — Ghost publish webhook to the ElevenLabs API, audio to your own storage, player injected into the post? I want the MP3 files themselves, since I also run a podcast and would rather generate once and use the file on both surfaces than have it locked inside a third-party player.
Roughly 27 posts in the back catalog, which I plan to do as a one-time manual batch. Automation is for go-forward posts only.