Own Voice: read-aloud for blogs, synthesised in the reader's browser

Listen buttons for blogs are a crowded category. ElevenLabs Audio Native, BeyondWords, TTS.ai and Listnr all serve Ghost or WordPress, and free WordPress plugins already read posts with the browser's built-in voices and highlighting. So a listen button on its own is not new. What I found no commercial product doing is running a neural model in the reader's browser, in the author's cloned voice, for a flat fee. The pieces exist only as open source: read-along and HeadTTS for an embeddable component with highlighting, and Pocket TTS, which OfflineTTS already runs in the browser with local voice cloning. A new entrant could win with small independent writers on two points. First, no per-character or per-listen cost, because the reader's device does the work, which is the main pain with ElevenLabs and BeyondWords. Second, privacy, since no article text goes to a third party. Pairing the widget with a local post-to-podcast tool in the same voice adds a second reason to pay; that space has only a weak cloud tool (LaunchPod), tutorials and a $49 automation template. The risks: model downloads are tens to hundreds of megabytes for each reader, Firefox audio problems were reported for in-browser Kokoro, Pocket TTS is English-only for now, and voice cloning needs real proof of consent. Also, incumbents or hobbyists can copy the core quickly because the models are open. Any lasting advantage would come from easy Ghost and WordPress setup, the voice-profile workflow and caching, not from the model. Documentation narration for open-source maintainers is unserved but has little budget, so it works better as a free channel than a paying market.

• ElevenLabs Audio Native

  Embeddable audio player with setup guides for Ghost and WordPress. It turns posts into audio on ElevenLabs' servers and supports cloned voices. This is the main paid incumbent for writers who want a listen button.

• BeyondWords (formerly SpeechKit)

  Listed in Ghost's integrations directory. It converts each post to audio on its servers and adds a customizable player at the top. Generation happens server-side, not in the reader's browser.

• TTS.ai Embed Widget

  One script tag adds a Listen button and lists Ghost as a supported platform. The page text is sent to a remote TTS API, and the free tier is limited by characters per request.

• Listnr Ghost integration

  Converts Ghost posts to audio you embed with an iframe, served from Listnr's CDN. Priced per word.

• Smart Audio Reader (WordPress plugin)

  Free plugin that uses the browser's built-in voices, so nothing goes to a server. It has a floating button and highlights the paragraph being read. This is the free floor a paid product has to clearly beat on voice quality.

• Audileo TTS (WordPress plugin)

  Another free read-aloud plugin using the browser's built-in Web Speech API, with a voice picker. No neural model and no author voice.

• LE-VAI/read-along

  MIT-licensed embeddable web component with word-by-word highlighting and swappable engines, including kokoro-js. It is the closest open-source version of the widget, but it's a developer component, not a hosted product, and has no voice cloning.

• HeadTTS

  Open-source Kokoro TTS that runs in the browser with WebGPU or WASM and returns word timings. It is an engine you could build on, not a drop-in listen button.

• OfflineTTS Pocket workspace

  Runs Kyutai Pocket TTS in the browser with ONNX Runtime Web and does voice cloning locally from a recorded or uploaded clip. Shows the core technique works today, but it is a standalone web app, not something you embed on a blog.

• Kyutai Pocket TTS

  The 100M-parameter English model with voice cloning that the idea relies on. Kyutai says it can run client-side in the browser. It is freely available to anyone, including competitors.

• LaunchPod

  Voice-cloning tool that narrates blog posts and makes podcast episodes in your voice. It launched on AppSumo with weak reviews. Cloud-based, overlaps with the post-to-podcast idea.

• freeCodeCamp blog-to-podcast pipeline

  Do-it-yourself tutorial that turns feeds into a podcast feed using a cloud voice-cloning API and Cloudflare R2. It shows demand for the CLI idea, but there's no local-model product.

Checked 2026-10-21