Podcastle is now Async
Async is the current name of Podcastle. The official January 28, 2026 announcement says the company changed its brand because the product had expanded beyond podcasting into video, code, audio, and developer infrastructure. Existing accounts, projects, pricing tiers, billing structures, and familiar workflows were carried forward. A current directory should therefore use Async as the canonical name and keep Podcastle as a historical alias and redirect, rather than publish a stale Podcastle page.
The platform still supports local-track remote recording for up to ten participants, audio and video editing, transcription, text-based edits, noise cleanup, leveling, filler-word removal, music, hosting, RSS publishing, AI voices, and Revoice cloning. The expanded product adds conversational editing, video and image generation, avatars, music and sound effects, clips, reframing, subtitles, thumbnails, dubbing, lip sync, upscaling, and developer voice APIs.
This breadth is valuable when it replaces several handoffs, but “all in one” is not evidence that every component is best in class. Recording reliability, transcript accuracy, edit precision, voice naturalness, dubbing, image and video quality, model licenses, turnaround, and unit cost should each have an acceptance test.
Recording, editing, and generation workflow
Remote sessions are locally recorded in separate participant tracks so an unstable connection does not directly determine the captured media. Still run a preflight for microphone, camera, permissions, disk, browser, headphones, lighting, backup recording, participant consent, and upload completion. Never close a session until every local track is confirmed on the server.
Text-based editing can remove words, silence, and filler from the media by editing the transcript. AI cleanup and auto-leveling accelerate routine work, but listen across every cut. Transcription errors can delete the wrong phrase, aggressive noise removal can damage voices, and automatic leveling can make intentional dynamics unnatural. Keep source media and a reversible project version.
Async now exposes more than 100 generation models across video, image, avatars, music, and sound effects. The price table identifies specific models and per-output credit costs, which is more useful than a single “AI credits” label. Select models deliberately: confirm provider, resolution, duration, audio, commercial terms, training opt-out, regional processing, safety policy, provenance, and whether a failed generation consumes credits.
For dubbing, correct the source transcript before translation, use a fluent reviewer, and check timing, lip sync, names, claims, tone, background audio, captions, and cultural fit. For voice or avatar cloning, retain signed permission that covers purposes, channels, duration, territories, editing, revocation, disclosure, model access, and deletion. Synthetic identity must never be used to deceive, authenticate, or imply an endorsement.
Plans, limits, and credits
The current pricing page lists Basic, Essentials, Pro, Teams, and custom Business. Its paid dollar amounts are dynamically rendered and were not reliably exposed in the official page text used for this review, so confirm the live monthly or annual checkout price. Async advertises a 40% saving on yearly plans.
Basic includes one creator, unlimited audio recording, three hours of video recording, 720p watermarked video, one hour of transcription, 10,000 TTS characters, 2 GB storage, and 10 lifetime AI credits. The current expanded plan table separately lists 450 monthly credits for Essentials, 1,200 for Pro, and 3,000 per Teams user, with fixed top-up packs; Business is custom. Model cost varies dramatically. A short image may cost one credit, while high-resolution video can use dozens or hundreds for a few seconds.
Build a project budget from approved deliverables, not raw generations. Include failed attempts, variations, extensions, upscaling, dubbing, transcription, storage, guest seats, exports, and human review. Check whether credits expire, renew per user, pool across teams, carry over, refund on failures, or survive downgrades.
Privacy, ownership, and security
The January 2026 privacy policy identifies Podcastle Inc. doing business as Async and defines voice recordings and derived voice models as biometric data. It offers access, portability, consent withdrawal, and immediate deletion requests for voice biometrics. That is useful, but buyers must still ask about actual retention, backups, subprocessors, model providers, storage regions, support access, training, and how deletion propagates into derivatives.
The terms say users retain ownership of input and output to the extent law permits and remain responsible for rights and lawful use. They also grant Async a perpetual, worldwide, royalty-free, sublicensable, irrevocable license to use input and output to provide the platform, conduct research and development, and improve Async and third-party AI services. Users may opt out of AI training by email. For confidential or client media, complete the opt-out before upload and obtain contractual confirmation; business and API accounts should rely on their MSA or API agreement, not assumption.
Async publishes a SOC 3 report covering a past audit period. Review the report date, scope, exceptions, complementary user controls, and whether the current post-rebrand services and third-party models are covered. Obtain the latest SOC 2 report under NDA where appropriate, DPA, subprocessor list, incident obligations, encryption and access details, business continuity evidence, and data-deletion terms.
Verdict
Async is a legitimate evolution of Podcastle and a strong candidate for teams that value an integrated creator pipeline. The rebrand is more than cosmetic: the platform now mixes established recording and podcast tools with a wide model marketplace and agentic editing. That makes it more capable and harder to evaluate.
Pilot one real production from recording to publication. Track upload success, edit time, transcription corrections, approved output per credit, generation failures, localization corrections, participant experience, rights paperwork, and final quality. Compare the result with a focused recorder/editor plus separate generation tools. Choose Async when the saved handoffs and shared workspace outweigh model-specific compromises and governance work.
