In short: Copy the article text, remove anything that only makes sense visually, paste it into Text to Speech in parts of up to 10,000 characters, choose a calm narration voice, generate, download the MP3 files and add them to your post with an audio player or a podcast feed.
Why add audio to your articles
Many people start an article and never finish it. An audio version keeps them with you when they cannot read: on a walk, in the car, or with tired eyes. It also helps readers with dyslexia or low vision, and gives you a second format to share as a podcast episode or a short clip on social media.
With AI narration there is no microphone, no studio and no re-recording when you edit the post. If you update a paragraph, you only regenerate that part.
Step by step
- Copy the text, not the pageCopy the article body only. Leave out menus, share buttons, image captions and "click here" links.
- Rewrite what does not work aloudTurn tables into sentences, write out abbreviations, and replace "see the chart below" with what the chart shows. Add the title and the author at the start.
- Split long postsEach conversion takes up to 10,000 characters. Split at headings so every part ends naturally; most posts need two to five parts.
- Pick a narration voiceCalm, clear voices work best for long listening. Try an HD voice or an audiobook or narration styled voice in your language.
- Add breathing roomLeave a new paragraph between sections and add
[break 1s]before each heading so listeners hear the structure. - Generate, check and downloadListen to each part once, fix any word that sounds wrong (spell it the way it is said), then download the MP3 files.
- PublishJoin the parts in any audio editor, then add the file to your post with an audio player, or upload it to your podcast host.
Which voice for which blog
| Type of content | Voice style that fits | Speed |
|---|---|---|
| How-to and tutorials | Clear, friendly narration or explainer voice | Normal |
| News and updates | News anchor or a crisp standard voice | Normal to slightly faster |
| Personal essays and stories | Warm audiobook or storyteller voice | Slightly slower |
| Business and B2B | Corporate or HD voice | Normal |
| Recipes and travel | Cheerful, conversational voice | Normal |
Common mistakes to avoid
- Pasting URLs. A voice reading "https colon slash slash" loses listeners. Remove links or say the site name.
- Leaving in code, tables or emoji. Describe them in words instead.
- Very long sentences. Split anything over about 25 words; it sounds more natural.
- Inconsistent voices. Use the same voice for every part and every post so your audio has a recognisable sound.
- Skipping the intro. Start with "You are listening to [post title] by [name]" so the file makes sense on its own.
Add a listen button to every post automatically
If you publish often, the Read Aloud widget adds a "Listen" button to your website that reads any page aloud, so you do not have to create files one by one. For long pieces such as reports and e-books, see document to audio and PDF to audio.
Frequently asked questions
Can I use the audio on my blog and podcast?
Yes. Audio you create can be used on your own website, podcast and social media. See the terms for details.
How long does a 1,500-word post take to narrate?
At a normal pace, about 10 minutes of audio. The voiceover length calculator gives an exact estimate.
My post is in Spanish. Does that work?
Yes. Pick a voice in the language of the post; there are voices in 75+ languages and accents.
How do I join several MP3 parts?
Any audio editor (Audacity is free) or video editor can place the parts one after another and export a single file.
Do I need an account?
You use your Speechma Pro account.
The Speechma newsletter
Make better voiceovers, faster
- New voices and features the day they launch
- Script templates and tricks that make AI voices sound human
- Creator stories, community picks and early access