How to Automate a Podcast with AI: Async (Formerly Podcastle) Tutorial for 2026

This tutorial takes you from an episode idea to a podcast published on Spotify and Apple Podcasts, then cut into short clips for social media, in eight steps with Async, the new name of Podcastle. AI drafts the outline, cleans up the audio, cuts filler words from the transcript, can stand in for your voice, and writes the episode summary. You need a browser (Chrome is recommended), a free account to try it, then the Essentials plan, at $19.99 a month or $11.99 a month billed annually, to export without the free tier’s limits (Async pricing, checked Oct. 22, 2026; dollar amounts archived Oct. 21, 2026).

Podcastle is now Async. The company renamed itself in January 2026, and podcastle.ai now redirects to async.com. It did not shut down: according to the official announcement, existing accounts, projects, pricing tiers, and billing carried over. The platform has added video, image and music generation, and a voice API; the help center calls the production app Async Creator. The steps below follow the Async help center as of Sept. 25, 2026. Screenshots from the March 2025 version of this guide, which showed the old Podcastle interface, have been removed.

Which AI Can Turn Text Into a Podcast?

Gemini Notebook, ElevenLabs GenFM, and Async can all turn text into a podcast without a mic, in two different ways. Gemini Notebook (the new name of NotebookLM) and GenFM, inside ElevenLabs Studio, write a conversation from your documents themselves, then have AI voices perform it. Async reads the script you give it, in the voices you pick, then lets you edit and publish the episode. Gemini Notebook can be used without a paid plan, GenFM requires a paid ElevenLabs plan, and Async’s free plan stops at about 15 minutes of text-to-speech in total.

Official sources checked Oct. 22, 2026: the Gemini Notebook help center (NotebookLM was renamed on July 16, 2026), the ElevenLabs help center and pricing page, and Async’s text-to-podcast page, pricing, help center, and terms (last modified Jan. 15, 2026). Async’s dollar prices come from its archived pricing data (Oct. 21, 2026). The ElevenLabs help center calls the feature GenFM; its button is labeled “Generate a podcast” in the how-to guide and “Create a podcast” in the Studio FAQ.

How to choose:

• Listening to your own documents (course notes, reports, research): Gemini Notebook, at no cost. The hosts summarize and discuss your sources; they don’t read your text word for word. More in our Google Gemini guide.
• Publishing a show based on an article or a book: GenFM, whose script you can fix before the audio is generated, with the commercial license that comes with the subscription.
• Controlling every sentence: Async, or an empty audio project in ElevenLabs Studio, reads your script as written. That’s the approach in steps 1 and 3 below.

In every case, you need the rights to the text you upload, and it’s good practice to tell listeners the voices are synthetic.

What You Need Before You Start

Sources: Async pricing, transcription languages, AI voices, and terms of service, checked Oct. 22, 2026; dollar amounts archived Oct. 21, 2026. The pricing page advertises 15+ languages for AI voices; the help center lists five.

How Much Does Async (Formerly Podcastle) Cost?

Async shows prices in the visitor’s local currency. The dollar prices below come from the data behind the official pricing page, archived on Oct. 21, 2026; from France on Oct. 22, 2026, the page showed the same figures in euros, unchanged since Sept. 25. Annual billing saves 40%: the yearly charge is $143.90 for Essentials, $287.90 for Pro, and $599.90 per creator for Teams, and the per-month figures in the table are those totals divided by 12, the way the page displays them.

Good to know before you pay:

• Two recording figures for Essentials: the plan card says 4 hours of audio and video recording per month; the comparison table says 2 hours of remote multitrack recording.
• AI credits pay for video, image, and music generation (Kling, Veo, Seedance, Nano Banana, and others), plus AI clips, subtitles, and dubbing. Magic Dust, noise reduction, and silence removal are included in every paid plan.
• Exports: 320 kbps MP3 or WAV from Essentials up, video in Full HD (Essentials) or 4K (Pro and Teams). The free plan caps video at 720p and adds a watermark.
• Free trial and refunds: annual plans come with a 7-day free trial; unless you cancel, the full year is charged when the trial ends and is no longer refundable (trials, refund policy). The pricing page notes that the plan’s AI credits unlock only after that payment. Without a trial, the help center and the pricing page promise a 7-day refund on a first subscription, and the pricing page excludes AI credits used; Async’s terms, last modified Jan. 15, 2026, say instead that no refunds are given once a charge has been processed. You can cancel anytime, with no prorated refund for the current period.

How to Automate a Podcast in 8 Steps

1. Draft the script with an AI assistant.
2. Record the episode, solo or with remote guests.
3. Or generate the voice from text, with a stock AI voice or your cloned voice.
4. Edit by transcript: filler words and silences.
5. Clean up the audio with Magic Dust.
6. Add music, an intro, and transitions.
7. Export and publish to Spotify and Apple Podcasts.
8. Repurpose the episode into short clips and show notes.

Step 1: Draft the Script with AI

A detailed outline is enough for a conversation; a full script works better for a read episode or one voiced by AI. An assistant like ChatGPT produces a first draft in seconds; fact-check it and rework it into your own voice. Our guide to writing high-quality content with ChatGPT covers prompting in more depth. A prompt to adapt:

You are a podcast producer. Write the outline for a 15-minute episode about [topic] for [target audience].
Structure: a 30-second hook, an introduction, 3 segments with one concrete example each, and a closing with a call to action.
For each segment, list 3 talking points and one question to ask the guest.
Write for the ear: short sentences, no jargon.

If an AI voice will read the episode (step 3), ask for a word-for-word script with numbers spelled out: text-to-speech reads them more reliably.

Step 2: Record the Episode, Solo or with Guests

Async records in the browser, with nothing to install. Each participant is recorded locally on a separate track, audio and video, and files upload while you record (Recording Studio page). The pricing page lists up to 10 participants per session.

Create a free account on Async commission.

To record with guests (official guide):

1. Log in, then click Create → Recording → Record with others.
2. Enter the recording title, your name, and your role.
3. Optional: click Customize to add your name, a logo, or a session title, then Save. You land in the waiting room.
4. Copy the Invite link for your guests, or invite them by email.
5. Once everyone has joined, click Record; when you’re done, click Stop Recording and end the session.
6. After the upload finishes, click Create audio project (or Create video project) to start editing.

Solo, go to Create → Recording → audio or video, pick your mic and camera, then Enter Studio and Record (recording guide). Each stop and restart is saved as a separate clip. A built-in teleprompter lets you read your script on camera.

Step 3: Or Generate the Voice Without a Mic

For a read episode (news roundup, summary, or a version in another language), text-to-speech replaces the recording (AI voices guide):

1. Create or open a project, then choose AI Voices in the left toolbar.
2. Filter by accent, tone, or category and preview the samples.
3. Click a voice, paste your text, then click Generate.
4. For a dialogue, click Add Paragraph and assign it another voice: each voice gets its own track.

Your own voice. In AI Voices, Create Voice builds a clone from a recording of up to 10 seconds or an uploaded file of up to 10 MB (voice cloning guide). Voice cloning requires Pro or higher (pricing, checked Oct. 22, 2026). Record somewhere quiet: the clone copies the sample’s flaws too. Only clone your own voice, or someone’s who has agreed to it.

With ElevenLabs. For more expressive voices, ElevenLabs commission generates the narration, which you then import into Async for editing. Its free plan gives about 10 minutes of speech per month for non-commercial use, and the Starter plan costs $6 a month with a commercial license and instant voice cloning (ElevenLabs pricing, checked Oct. 22, 2026). Its Studio also offers GenFM, which writes and voices a podcast from a document, for paid subscribers only (comparison above). More in our ElevenLabs guide and our guide to creating realistic AI voices with ElevenLabs.

Step 4: Edit by Transcript (Filler Words and Silences)

Text Mode in the audio editor shows the transcript of each track, so you cut in the text instead of scrubbing the waveform (Text Mode guide).

1. In the audio editor, switch to Text Mode, select the tracks, set the language, and click Transcribe.
2. Filler finds filler words (um, uh, mm, mhm, and similar): Remove all deletes every one, or remove them individually with Remove (guide).
3. Silences finds pauses longer than 0.5 seconds (adjustable) with the same removal options (guide).
4. To cut a passage, click a word and choose Remove: it disappears from both the transcript and the audio. Correct, on the other hand, only fixes the transcript, not the sound (Words function guide).

For a video episode, Rough Cut (in the Create menu, under video) removes filler words, retakes, and unnecessary silences in one pass; the video editor’s AI Chat can also run it on request (Rough Cut guide).

Step 5: Clean Up the Audio with Magic Dust

Magic Dust AI reduces background noise (air conditioning, keyboard clicks, traffic), balances frequencies, and smooths the voice (Magic Dust guide):

• one clip: right-click the clip, then Add Magic Dust AI;
• several clips: select them, then click the Magic Dust AI button in the toolbar;
• the whole project: open AI Assistant in the bottom-right corner, then Apply Magic Dust AI.

A slider then sets the intensity. Magic Dust already includes noise reduction, Auto Leveling, and silence removal, so Async advises against stacking those on top of it. It is built for voice, not music.

Step 6: Add Music, an Intro, and Transitions

The audio editor includes a royalty-free library of more than 7,000 music tracks and sound effects (music and SFX guide):

1. Open the project’s Music & SFX folder.
2. Search by genre, mood, instrument, or effect (“upbeat,” “whoosh”) and preview.
3. Drag the track onto the timeline or click + Add to track.
4. Add a fade-in and fade-out so the intro music ducks under your voice.

The free plan only unlocks part of the library. If you also publish on YouTube, paste the track’s license code into the video description. For a theme that’s yours alone, see our guide to composing music with AI, then import the file into your project.

Step 7: Export and Publish to Spotify and Apple Podcasts

Export. The Export button in the top-right corner offers MP3, WAV, FLAC, AAC, and other formats. Turn on the transcript options to get the transcript and an episode summary (Include episode summary) along with the audio, handy for show notes (export guide).

Host on Async. The Hosting Hub, billed as free with unlimited episodes (Hosting Hub page), gives you an RSS feed and a show page:

1. Create the show: Shows → Create a show, then add the title, description, author, language, categories, explicit-content setting, and square cover art between 1400 and 3000 pixels; click Finish (guide).
2. Publish the episode: from the project, click Export → Publish and pick the show; add episode details and artwork, then publish now or schedule it (guide).
3. Distribute: copy the RSS link in the top-right corner of the show page and submit it once to Spotify for Creators (which emails a verification code) and to Apple Podcasts (Spotify guide).

From then on, every episode you publish on Async is added to the feed and reaches those platforms automatically (distribution guide). If you already have a podcast host, just upload the exported file there.

Step 8: Repurpose the Episode into Clips and Show Notes

AI Clips turns a video between 3 minutes and 2 hours long (up to 8 GB, MP4, MOV, or AVI) into captioned vertical clips for TikTok, Reels, and Shorts. Start it from the home page, choose the language, the aspect ratio (9:16 by default, or 1:1, 4:3, 16:9), and the caption style; you get an email when the clips are ready (AI Clips guide). AI subtitles are English-only on Essentials and Pro, and cover 5+ languages on Teams (pricing, checked Oct. 22, 2026).

For promotion, give the exported transcript to your AI assistant:

Here is the transcript of a podcast episode. Write:
1. an episode title of 60 characters or fewer;
2. a 150-word description for Spotify and Apple Podcasts covering the 3 key takeaways;
3. timestamped chapters based on the transcript's timestamps;
4. three posts for LinkedIn, Instagram, and X, each with an exact quote from the episode.
Don't add anything that isn't in the transcript.

[paste the transcript]

To go further, an automation tool like Zapier can watch your RSS feed and trigger a social post whenever a new episode goes live.

Common Mistakes and Fixes

Some Filler Words Survive

The default Filler list covers sounds like “um,” “uh,” “mm,” and “mhm,” not verbal tics such as “like” or “you know.” Find those with the search icon in Text Mode, then Remove all, and listen back: cuts that are too tight make the delivery sound choppy.

Fixing the Text Doesn’t Change the Audio

Correct only edits the transcript. To cut audio, use Remove.

The Audio Sounds Worse After Processing

According to Async, Magic Dust on a track already processed with Noise Reduction or Auto Leveling, or on music, can distort the result. Undo (Ctrl + Z) and keep Magic Dust alone, on voice tracks only.

A Guest’s Track Is Missing

If a participant’s upload fails, a message in your files flags it with a recovery link: try Retry first, then send the link to the guest so they can finish the upload. Before the session, tell guests to use Chrome, outside incognito mode, with only one Async tab open.

The Free Plan Runs Out After One Episode

Its quotas (1 hour of recording, 2 hours of transcription) apply over the life of the account, not per month. It’s for trying the tool; a regular show needs Essentials. Download your files in time, too: on a free account, every file you upload, record, or generate (audio, video, transcripts) is permanently deleted 30 days after it’s uploaded or created (terms of service).

Async Alternatives for Automating a Podcast

Async puts everything in one place. Other tools match or beat it on specific steps (prices checked Sept. 25, 2026 on the official pages):

• Riverside fits shows where guest recording quality comes first; separate-track downloads are capped at 15 hours per month on Pro.
• Descript fits creators who mostly edit themselves, working from the transcript.
• Adobe Podcast is enough to rescue poor audio for free; its Premium tier now comes with the Adobe Firefly Pro plan.
• ElevenLabs pairs with any of the three for synthetic voices and translated versions.

FAQ

Does Podcastle Still Exist?

Yes, as Async since January 2026. podcastle.ai redirects to async.com, and existing accounts, projects, and subscriptions carried over (official announcement).

Is Async (Formerly Podcastle) Free?

There is a free plan with no credit card required, but its quotas don’t renew: 1 hour of recording, 2 hours of transcription, about 15 minutes of text-to-speech, and a single exported file. Files are deleted 30 days after upload or creation. The first paid plan is Essentials, at $19.99 a month or $143.90 a year, with a 7-day free trial on annual billing (pricing, checked Oct. 22, 2026; dollar amounts archived Oct. 21, 2026).

Can I Make a Podcast Without Recording My Voice?

Yes. AI voices from Async or ElevenLabs can read your script, and voice cloning lets you use your own voice without recording every episode, on Async Pro or from ElevenLabs Starter. Gemini Notebook and ElevenLabs GenFM go further: they write the conversation from your documents themselves (see the comparison). Tell listeners when a voice is synthetic.

Can I Host My Podcast on Async for Free?

The Hosting Hub is billed as free, with an RSS feed, a show page, and unlimited episodes. You still need to submit the feed once to Spotify and Apple Podcasts.

Async or Riverside: Which Is Better for Podcasts?

Both record remotely, transcribe, edit by transcript, and host. Async adds text-to-speech, voice cloning, and image and video generation; Riverside focuses on recording, live streaming, and webinars. Riverside Pro costs $29 a month when billed monthly (checked Sept. 25, 2026), versus $19.99 for Async Essentials (archived pricing data, Oct. 21, 2026).