Async formerly known as Podcastle is the easiest way I know to start a podcast without installing anything or spending a cent upfront. You record, clean up, transcribe, and export entirely in the browser, which removes the usual barrier of stitching together a DAW, a noise-reduction plugin, a transcription service, and a separate remote-recording tool. I point new podcasters here when the goal is to get a show out the door rather than chase studio-grade production. For a solo creator, that consolidation is worth more than a few decibels of extra fidelity.
How the recording workflow actually goes
You open a recording project, pick local or remote, and hit record. For a solo episode you talk into your mic and the browser captures the track. For an interview you send guests a link, they join from their own browser with no download, and each person records on a separate track so you can balance levels afterward. That separate-track capture is the part that saves you later, because a guest who is too quiet or who coughs into the mic can be fixed without touching your own audio. I run a 30-second test first every time, watch the level meter, and aim for a peak around -12dB before recording anything I plan to keep. That one habit prevents the most common rookie problem, a track recorded so hot it clips or so soft that enhancement amplifies the room hiss along with the voice.
What the AI editing tools handle
The cleanup tools are where Async earns its place for people who hate editing. Silence Removal and Filler Word Removal both run in one click with solid accuracy, and that single pass clears out most of the dead air and the "um" and "you know" padding that otherwise eats an hour per episode. I treat them as the first edit pass before any manual trimming, because they tighten the timeline and leave a much shorter thing to refine by hand. Magic Dust is the enhancement AI that does the actual sound improvement, running compression and noise reduction internally to lift a voice out of a noisy room. The catch is that Magic Dust is already doing a lot, so stacking your own EQ or a second compressor on top of it muddies the result. Let it do its job and resist the urge to over-process.
Voice cloning and transcription
Revoice is the voice-cloning feature, and it is the most ambitious thing in the suite. You record a 70-sentence training session, and Async builds a clone you can use to generate AI-read audio for the moments you cannot get to a mic, like fixing a flubbed sentence or reading a late-breaking sponsor mention. It works, but quality varies a lot by voice type, so treat your first clone as an experiment rather than a guaranteed match. Some voices come back convincing and some land in an uncanny middle ground. Transcription runs alongside recording and gives you a text version of the episode for show notes, blog repurposing, or accessibility. The transcript is good enough to publish after a light proofread.
Pricing and what you actually get
The free tier is genuinely usable for a new show, with unlimited recording, basic AI audio enhancement, remote guest recording, and no watermark. That is more than most free podcast tools give you. The Essentials plan at $19.99/month, or $11.99/month billed annually, is the main upgrade and adds higher-quality exports, unlimited AI enhancement, and Revoice voice cloning. A Pro tier above that runs higher for heavier creators who need more hosting and download headroom. The honest read is that the free plan gets you launched and Essentials is what you move to once export quality and unlimited enhancement start mattering, usually around the point you commit to a regular publishing schedule.
Where it falls short
The audio quality ceiling sits below Riverside for remote multi-guest recordings. Riverside's local-recording architecture captures noticeably cleaner guest audio, so if your show lives or dies on interview sound, that gap is real. Revoice asks for the full 70-sentence training session and the clone quality is hit or miss depending on the voice, so do not build a workflow around it until you have tested your own. Export options are thin on the free plan, with MP3 quality capped below studio standard. None of these are dealbreakers for a solo show, but they are the exact places where a more specialized tool pulls ahead.
Who it's for and who should skip it
Solo podcasters and low-footprint interview shows that value getting started over squeezing out the last bit of fidelity are the core fit. If your guest audio is secondary and you want one tool for the whole pipeline, Async covers it from record to export. Skip it if you produce video podcasts or if pristine remote guest audio is non-negotiable. In those cases Riverside is the upgrade worth paying for once quality becomes the bottleneck. Skip it too if you already own a DAW you are fluent in and only need one missing piece, since Async is built to replace the whole stack rather than slot into it.
Getting the most out of it
Set your recording level once during the test, aiming for about a -12dB peak, then let Magic Dust handle the cleanup. Do not stack additional EQ or compression on top of it. Run Silence and Filler Word Removal as your first edit pass before any manual trimming, since they clear out most of the dead weight and leave you a tighter timeline. For interviews, lean on the separate-track recording and fix guest levels individually rather than processing the whole mix at once. And generate the transcript early so your show notes are ready the moment the audio is.
