Turn a Blog Post into a Two-Host Podcast (Without It Sounding Like a Reading)

A 1,500-word blog post becomes roughly a 10-minute, two-host episode for about 950 points if you let the podcast studio write the dialogue, or about 600 if you write the script yourself and only pay for the voices. The audio is the easy part: a 10-minute episode runs about 600 points of voicing, and nothing else is charged. Making two voices sound like a conversation, not a post read in alternating paragraphs, is the work.

This guide covers the two ways to start from a post, how to edit the generated dialogue, how to cast two voices that don't blur together, and what the whole thing costs. If you want to hear a first exchange before reading on, the two-host podcast generator writes one from a single sentence.

How to turn a blog post into a podcast

At a speaking pace of about 150 words a minute, a 1,500-word post runs 10 minutes, and a 3,000-word one runs 20. Podcast Studio takes either a brief or a finished script. Which one you use decides how much of the post survives into the episode, and how much of the thinking you do yourself.

• A. Brief it and let the studio write — State the post's central question as the topic, name two speakers with different roles, pick a format and a length. The studio writes the script and attributes every line. It costs points, and the price is shown before you commit.
• B. Write the dialogue yourself — Rework the post into lines labelled NAME: text and paste them in. Speakers are detected automatically and no AI is involved until you generate audio, so this half is free.

Route A is faster. Route B is better when the post is your own and you care about specific phrasings, statistics or a product claim being exact. A hybrid works well: have the studio write the first draft, then edit it by hand, which is where the rest of this post comes in.

One thing to be clear about: if you simply want one voice to read the post, with a title read, the body and a closing line, that is a different tool. The blog post to podcast episode page covers it, and a single narrator is often right for tutorials. Choose two hosts when the post makes an argument that someone could push back on.

Pick a format for a two-host podcast from an article

The topic is the biggest lever on quality, and a question beats a field. Remote work is a field. Whether remote work made engineering teams more productive, or just moved the meetings into chat, is a question. A post that already has a thesis usually has the question inside it.

• Co-hosts

  Two people who both know the material. Best for an explainer or how-to, and the safe default.

• Interview

  A curious host and one expert. Best for a deep dive where one person carries the substance.

• Debate

  Two opposed speakers who end still disagreeing. Best for an opinion piece or a versus comparison.

• Panel

  A moderator and up to 3 guests with distinct angles. Best for a roundup post.

If you're unsure, pick Co-hosts. There's also a Heat setting, from Civil to Combative, for how hard speakers push back. Engaged suits most posts, since it means real disagreement argued in good faith.

Edit the dialogue so it doesn't sound read

The most common failure is two hosts who politely hand a lecture back and forth. Every turn is a tidy paragraph, nobody interrupts, and the second host only ever says "Great point." Fix that on the page before you spend points on audio.

Notice the moves. The turns are short, one host asks what a reader would ask, and the second line pushes back instead of agreeing. Here's the checklist I'd run on any generated script.

• Split long turns. Anything over about 60 words, roughly 25 seconds of speech, becomes 2 turns with a short reaction between them.
• Add one-word reactions. A "Right." or "Mm-hm." between two long turns does more for realism than anything else you can do by hand.
• Cut one line off mid-thought. A trailing dash on an interrupted sentence is the closest you can get to overlap, because speakers can't talk over each other.
• Delete the post's signposting. "As mentioned above" and "In this section" belong on a page.
• Make the second host disagree once. One honest pushback early shapes the whole episode.

In the editor you can reorder lines, split a turn, add a pause between speakers and set an emotion or a delivery note on a single line. Tighten the gaps: podcast turns run closer together than audiobook narration, and short pauses make the exchange feel live. If you enable laughs and reactions in the brief, a [laugh] or [sigh] is allowed in the script, and a little goes a long way.

Cast two voices that contrast

Two warm, enthusiastic hosts are indistinguishable. The fix is contrast, in the voices and in the personalities you write. Give one host a clear, measured voice and the other a friendlier, quicker one. Use different genders or ages if you can, and let one speaker's personality be blunt where the other is warm.

Auto-cast will fill every empty seat in one click, and you can swap any host for a different voice afterwards without touching the script. Assigning a voice once re-voices all of that speaker's lines. Here are two that pair well. For more on casting a show host, see the AI podcast voice generator guide.

• 

  Confident Female Narrator Voice

  The host who asks and steers

  • American
  • Confident, friendly
  Use in Studio
• 

  Lucid Male Explainer Voice

  The co-host who explains in detail

  • American
  • Clear, measured
  Use in Studio

Length and cost in points

A podcast is charged in two parts. Writing the script costs 200 points plus 15 per minute of episode. Voicing it costs about 1 point per second of audio, the same as any other generation, and you're charged per line as you generate. An episode you never voice only costs you the writing.

Put those against the plans. Lite's 3,000 points covers three 10-minute episodes with the studio writing them, or five if you paste your own scripts. Pro's 12,000 covers twelve. A target length is a target, not a promise, so ask for a little under what you need and extend in the editor, which costs nothing.

Save points by generating segment by segment. Listen to the first section before voicing the rest, and regenerate single lines instead of whole episodes. One limit to plan around: speakers can't talk over each other, and music beds and intro stings aren't available yet.

Frequently asked questions