Support

Telltake is made by one person. Reports are read; replies may take a few days.

Report a problem

The diagnostics report. In Telltake choose Help → Save Diagnostics… and attach the file. It contains error codes and where in the app they happened, plus your macOS and app version. It contains no transcript text, file paths, recordings or provider responses, so it is safe to share.
What you did and what you expected. Which button, on which window, and what happened instead. If a message appeared, copy the code in square brackets at its end, for example [export.run · LocalExportError.sourceUnavailable].
About the recording, in words. Number of clips, roughly how long, camera or phone model, and whether it is HDR. Do not attach the recording itself unless asked.

A failed cloud request is never retried on its own and is not charged twice by Telltake; check your provider dashboard for what was billed.

"Apple could not verify Telltake" on first launch

Telltake is signed, but not with a certificate from Apple's paid developer program, so macOS cannot check it against Apple's notarization service and shows a warning. The app is unchanged by this; it is a statement about the certificate, not about the software. To open it once:

macOS 15 or later: dismiss the warning, open System Settings → Privacy & Security, scroll to the bottom, click Open Anyway, and confirm with your password or Touch ID.
macOS 13 or 14: right-click (or Control-click) Telltake in Applications, choose Open, then click Open in the dialog.

After that, Telltake opens normally. If you would rather not do this, the alternative is to build the app from source on your own Mac, which produces the same warning-free result because your Mac made it.

Keys and cloud processing

Doubao key: create a speech API key in the Volcengine console with the recording-file recognition model enabled, then paste it under Cloud settings… in Telltake. It is checked the first time you transcribe.
OpenAI key: paste it under Cloud settings and choose Validate key…. Validation only checks that the key can use GPT-5.6 Sol; it sends no recording or transcript. If validation says the model is not available to your key, enable it for that project in the OpenAI dashboard.
Stopping: Revoke cloud processing blocks new uploads and tries to cancel requests in progress. Remove key deletes a key from your Keychain. Details on the privacy page.

Known limitations in 1.0

One speaker and one camera, primarily Mandarin with natural Mandarin–English mixing. Multi-speaker recordings and dialect-heavy speech are outside what this version is for.
The interface is English. A Simplified Chinese version is planned.
Edits keep the recording's order. There is no timeline rearrangement, no captions, music, effects or publishing.
The AI Story Cut (a shorter proposal for a target length) is not in this version.
A sequence cannot mix HDR and SDR clips. HDR clips export as 10-bit HEVC with the source's colour tags; Dolby Vision dynamic metadata is not carried.
Cut quality and time saved are targets, not measured results. The app shows every removal and lets you restore it; review before you export.
Telltake is not sandboxed and needs no special permissions, but it reads only the clips you add and writes only where you choose.

Where things live on your Mac

Projects are .easycut packages wherever you saved them. They reference your clips by path and fingerprint; if you move a clip, use Relink selected clip….
Back up project… copies the project data only; keep your original clips separately.
Keys and your cloud agreement are in the login Keychain; a revocation marker is in ~/Library/Application Support/Easycut/.