Async Review

In January 2026, Podcastle stopped being a podcasting company. It rebranded as Async and repositioned around a much broader claim: one platform where a creator edits video, a marketing team localises content, and a developer builds a voice agent, all on the same models.

That is a substantial change of scope, and it is worth understanding before evaluating anything the product does today. Anyone arriving from a bookmark, an old tutorial or a two-year-old recommendation is looking for a tool that no longer exists under that name.

The underlying technology did not change. What changed is who it is sold to and what the interface asks you to do — which is now, primarily, to describe the edit rather than perform it.

Async, Formerly Podcastle

• Podcastle rebranded as Async in January 2026, moving from a podcasting tool to a broader AI media platform.
• Existing accounts, projects, pricing tiers and billing carried across unchanged, per the company's own announcement.
• Two divisions: Async Studio for creators, Async API for developers building on the same models.
• Editing happens through conversation — fix rough cuts, cut to the beat, clean audio, generate B-roll, add captions.
• Magic Dust is the audio cleanup, alongside dubbing, lipsync, voice cloning, eye contact correction and background blur.

What Did The Podcastle Rebrand Change?

The company announced the rebrand in January 2026, framing it as a shift from a dedicated podcasting tool to a unified AI media platform, with the CEO's stated position that the next generation of creators want to direct rather than edit.

Practically, three things changed. The name and domain moved to async.com. The scope widened from audio-led podcast production to video as the primary surface. And the business split into two: Async Studio for people making content, Async API for developers embedding the same capabilities.

What did not change matters as much. The announcement states that accounts, projects, pricing tiers and billing structures all remain the same, and that users keep access to their existing workflows. That is the right way to handle a rebrand, and it is not always how they are handled.

The company reports $23.5M in total funding, with backers including Andrew Ng's AI Fund, Mosaic Ventures, RTP Global, Point Nine and Sierra Ventures. That is relevant mainly as a signal about durability — a rebrand this broad is usually a bet funded by someone.

The honest question for an existing Podcastle user is whether a broader product serves them better. A podcaster who wanted a podcast tool now has a video platform with podcasting inside it, and breadth frequently means the original use case stops getting the attention it did.

Renames are more common in this category than they should be, and each one quietly breaks documentation. A tool such as quso.ai went through the same thing, and so did Gemini Notebook.

The announcement is explicit that accounts, projects, pricing and billing carried across unchanged. That is the right way to handle a rebrand, and not how they are usually handled.

Studio And API Are Two Businesses

Async operates through Async Studio and Async API, serving distinct users on the same models, and that structure explains most of the product decisions.

Studio is the creator-facing application: record, edit, clean, caption, clip, publish. It is judged on whether the interface makes a finished video faster.

API is infrastructure: the same speech synthesis, recognition and voice models, called from someone else's software. It is judged on latency, reliability and cost per unit, and its customers never open the Studio.

Running both from one model stack is efficient and creates a genuine tension. The features an API customer wants — determinism, parameters, predictable output — are frequently the opposite of what a Studio user wants, which is a good default and no decisions.

For a buyer, the useful read is which side you are on. If you are choosing between Async and a creator tool, compare Studio and ignore the API entirely; if you are building a product, compare the API against speech infrastructure vendors, where the competition is quite different.

Chat-Driven Editing Is Faster And Coarser

Async's central interaction is conversational: you describe the edit — remove the filler, cut to the beat, clean the audio, generate B-roll — rather than performing it on a timeline.

For coarse work this is genuinely faster, and the reason is that describing an intention takes seconds while executing it takes minutes. "Remove every pause longer than a second" is one sentence and a hundred manual edits.

It is also better for people who cannot edit. A timeline is an expert interface, and a large number of people who need video produced have no intention of learning one — which is exactly the market a chat interface opens.

For precise work it is worse, and predictably so. Any edit where the exact frame matters, where the decision is aesthetic rather than describable, or where you would know it when you see it, is a job for a timeline. You cannot describe a cut you have not yet chosen.

The workable pattern is therefore chat for the pass that removes obvious problems, and a timeline for anything the result still gets wrong. Treating conversational editing as a first pass rather than as the whole workflow is what keeps it useful.

The nearest philosophical neighbour is Descript, which edits the transcript rather than the conversation — both are attempts to escape the timeline, from different directions.

• Use conversational editing for the coarse pass — filler, pauses, bad takes, levels.
• Move to a timeline for anything where the exact frame matters.
• Check dubbed output with a native speaker before publishing to that market.
• Get written consent for any cloned voice that is not your own.
• Compare Studio if you are a creator, and the API separately if you are building.

Magic Dust And The Audio Half

Magic Dust is Async's audio cleanup — noise removal and voice clarity — and audio remains where the platform's heritage shows.

That heritage is an advantage. Podcastle spent years on speech, and speech is the hardest part of most video: viewers forgive poor footage far more readily than poor sound, and bad audio is the most common reason an otherwise decent video is unwatchable.

The cleanup tools are also the ones with the clearest limits. Noise reduction attenuates what an algorithm believes is noise, and pushing it hard produces the watery quality that identifies over-processed speech instantly. The right target is not distracting rather than clean.

Voice cloning, dubbing and lipsync sit alongside, and these carry an obligation rather than a limitation. Cloning a voice that is not yours, or dubbing a guest into another language in their own voice, requires that person's explicit written consent for that specific use — a release for recording an interview does not cover synthesising new speech from it.

Eye contact correction is the one to use most sparingly. It works by altering where someone appears to be looking, which is a small change with a strange effect: viewers frequently sense that something is off without being able to identify what.

For audio that needs rescuing rather than cleaning, none of this is enough, and a spectral repair suite like iZotope RX is the tool that actually recovers a damaged take.

Moving A Podcast Workflow Onto Async

For an existing Podcastle user, the practical question is whether the same workflow still runs. For a new user, it is whether a video-led platform suits an audio-led job.

Either way, test on an episode you have already published, where you know what the finished version should sound like.

1. Take a published episode and run the raw recording through the conversational cleanup pass.
2. Compare the result against the version you actually shipped, listening on laptop speakers rather than headphones.
3. Check specifically what the automatic filler removal took out that you would have kept — pauses carry meaning in interviews.
4. Test one dubbed language with a native speaker before assuming localisation is solved.
5. Verify that your existing projects and settings carried across the rebrand as stated.
6. Decide separately whether the video features are useful to you, or whether you now pay for a platform you use a quarter of.

Async: Rebrand And Backing