Source preserved locally
The saved recording remains separate from transcript and Summary attempts.
Native Mac voice memory
RecordCove saves authorized meetings and voice memos as local source audio, then lets you transcribe, review, organize, and export without a silent cloud fallback.
Public download is not available yet.
Source audio remains independent from every transcript attempt.
The saved recording remains separate from transcript and Summary attempts.
The current transcription path runs with installed local models or reports that it is unavailable.
Recording, transcription, review, and organization do not require a RecordCove account.
Source before transcript
RecordCove keeps the saved source audio independent from transcription, editing, and understanding. If a transcript attempt fails, the retained source remains available for another supported attempt.
How it works
Each stage is explicit, and the source remains separate from the results derived from it.
Capture authorized audio with the Mac app, or import one supported CAF, WAV, or M4A file.
RecordCove completes the local source before treating any transcript as the result.
Choose an installed transcription model. RecordCove records each attempt and never falls back silently to a remote model.
Edit the transcript or Summary, compare attempts, add categories and tags, then copy, export, or back up deliberately.
Current product scope
RecordCove keeps the local source, derived text, and user-controlled copies as distinct records.
Record authorized meetings and voice memos with the native Mac app.
Microphone permission is separate from participant consent.
Import one CAF, WAV, or M4A recording and retain an exact local source copy.
Bulk import, video, URLs, and drag and drop are not part of the current claim.
Run an installed local model, retry, or switch models while preserving earlier attempts.
AI output remains provisional and should be reviewed.
Edit transcript versions, titles, and structured Summaries without rewriting earlier history.
Summary quality depends on the selected transcript and available local Understanding path.
Search visible recording content, sort and filter the library, and manage categories and tags.
Speaker diarization and speaker search are not available yet.
Keep simple local text notes with titles, search, sorting, and a focused writing view.
Quick Notes are separate from recordings and model processing; recording-linked annotations are not available yet.
Restore deleted recordings and Quick Notes for 30 days, or confirm permanent deletion separately.
Deleting inside RecordCove does not delete earlier backups or exports.
Copy or export text as TXT, Markdown, CSV, or PDF and audio as Original, WAV, or M4A.
Exports and backups are unencrypted external copies that remain outside RecordCove.
Who it is for
The early product is designed for personal, authorized capture and review. It is not a team workspace or compliance-grade transcription service.
Keep authorized calls, decisions, and follow-up context available for later review.
Revisit client-approved sessions without making remote processing a hidden dependency.
Review authorized interview recordings and provisional transcripts in one local library.
Organize consented interviews and field recordings with categories, tags, and exports.
Privacy and control
The current native workflow keeps recording, playback, transcription, attempt history, organization, and available Understanding processing on the Mac. It does not silently send audio or transcripts to a remote model.
Downloading a local model is a user-initiated network action. Using that installed model locally does not upload the recording or transcript merely because another organization trained it.
Deleting a recording or Quick Note moves it to Recently Deleted for 30 days. You can restore it during that period or confirm permanent deletion separately. Existing backups and exports are separate files and are not deleted automatically.
Backups and exports are unencrypted files. The destination owner controls their access, protection, retention, and deletion.
Microphone permission allows capture on the Mac. You must still record only when authorized and obtain participant consent where required.
Local models
RecordCove distinguishes the model that turns audio into text from the local path that creates a title and structured Summary.
Transcription model
Choose an installed Whisper or Parakeet model for a transcript attempt. The attempt records the requested and actual model details for later review.
Public model availability, language scope, redistribution rights, speed, and accuracy remain subject to the final release matrix.
Understanding model
Use Apple's on-device model where the Mac and locale are eligible. If it is unavailable, RecordCove preserves the transcript instead of substituting an unqualified processor.
Only Understanding models that pass the executable release gate may appear as selectable choices.
Requirements
The current development target is an Apple Silicon Mac running macOS 14 or later. Final public support, memory, disk, microphone, accessibility, language, and model-storage requirements are still being verified against the release package.
FAQ
Not in the current local workflow. Installed transcription models run on the Mac, and RecordCove does not silently fall back to a remote model.
Only for explicit or future external actions. Downloading a local model is an explicit network action. Public feedback, updates, and remote processing are not active on this pre-launch page.
The transcript attempt records its failure without deleting the retained source audio. You can retry with a supported installed model.
No. RecordCove can inspect and repair supported interrupted recordings while retaining the original, but it does not promise recovery from every form of corruption or device failure.
Current native export supports transcript and Summary copies in TXT, Markdown, CSV, or PDF, plus audio as Original, WAV, or M4A. Exported files are unencrypted and live outside RecordCove.
No. Speaker diarization and voice labeling remain future work and are not part of the current product claim.
Beta terms and pricing are not published. RecordCove will not present pricing until the offer, limits, support, privacy, and release terms are approved.
No public download is available yet. Signing, notarization, package verification, Homebrew, support, and release review must pass first.
Public beta status
RecordCove is completing signing, notarization, package verification, support, and release review. Download and Homebrew actions will remain disabled until they point to the same independently approved artifact.
Feedback: No public RecordCove feedback destination has passed verification yet, so this page does not link to one.
Updates: The current manual action reports that updates are unavailable and does not contact a server. Production discovery, verification, download, and installation are not active.