Harvestry Documentation FAQ

Frequently Asked Questions

Everything people ask most often — buying and licensing, languages and translation, privacy, compatibility, and troubleshooting.

Buying & Licensing

Does Harvestry require a subscription?

No. Harvestry is a one-time purchase. You pay once and receive all updates within that major version (v1.x) at no extra cost. Future major versions — when they arrive — will be offered as a new purchase, with a meaningful discount for existing customers.

Can I try it before buying?

Yes. Harvestry includes a 7-day free trial — no account or credit card required. Download it, process a few of your own videos, and see the output before you spend anything. The trial is deliberately long enough to evaluate Harvestry against real content.

What's the difference between Standard and Pro?

The core pipeline is identical, and transcription in every language Whisper supports is included in both. Pro adds three things for users who want to go deeper:

If none of those matters to you, Standard is the right choice. See Standard & Pro for a full comparison.

How many Macs can I use Harvestry on?

Each licence covers up to 2 Macs. When you're signed in to the same iCloud account on both, your library (lectures, transcripts, screenshots, consolidated notes, and annotations) syncs between them automatically. Exported study documents sync too, as long as Save exports to iCloud Drive is on. Video files don't sync; each stays on the Mac that downloaded or imported it.

What is your refund policy?

All sales are final. Because Harvestry includes a full 7-day trial with no credit card required, we encourage you to try it thoroughly before purchasing. If you encounter a technical issue that prevents the app from working on your machine, contact us at support@harvestry.co and we'll work to resolve it — and if we can't, we'll make it right.

Languages & Translation

Which languages does Harvestry handle?

Whisper covers 100 languages, and Harvestry transcribes each one in the language it was spoken — French stays French, Japanese stays Japanese. That is included in both Standard and Pro, as is choosing the spoken language yourself rather than relying on auto-detection.

Which languages can Harvestry translate?

Translation goes into English from any of the other 99. It cannot translate into French, German, or anything else — that is a limit of the speech model, which was trained to translate into English and no other target, rather than a Harvestry restriction.

Quality also varies considerably by source language. Widely spoken languages — French, Spanish, German, Italian, Portuguese, Dutch, Japanese and their neighbours — are excellent. Low-resource languages such as Maltese, Amharic, Shona, Lao, and Yoruba are supported but noticeably weaker, and the failure mode is fluent-sounding English that doesn't match what was said rather than obvious nonsense. Everything runs locally and re-running costs nothing, so test your own material during the free trial before buying.

Can I get notes in a language other than English?

Yes — through consolidation rather than transcription. This is a Pro feature, and it's the one to reach for whenever the language you want isn't English.

  1. Transcribe in the spoken language (leave the Transcription menu on Transcribe, not Translate).
  2. Set Settings → Consolidation → Notes language to your target.
  3. Choose Reprocess → Run consolidation on the lecture.

The consolidation model writes the notes in your chosen language, so pairs like French → German work here even though the transcription stage can't produce them. Quality depends on the model: large hosted models handle most languages well, while small local Ollama models may struggle with less common targets. See Notes Language for a worked example.

What about the app's own interface?

Harvestry's interface is localised into English, German, Italian, Spanish, and French, and follows your macOS language preference automatically. This is unrelated to what it can transcribe — your videos can be in any of the 100 supported languages whichever interface language you use.

Consolidation & Privacy

Do I need an Anthropic or OpenAI API key?

Only if you want to use the optional LLM consolidation step with Claude or ChatGPT. The core pipeline — transcription, translation, screenshot capture, and HTML export — works entirely offline with no API key of any kind. Pro users can also run consolidation locally through Ollama, which needs no key either.

Note that an OpenAI API key is billed separately from a ChatGPT Plus subscription; a Plus subscription alone does not provide API access.

Does any of my content leave my Mac?

All processing happens on your Mac: transcription, translation, screenshot capture, and building the export never use a server, and nothing is ever sent to us. Data can leave your Mac in two ways:

Can I stop my library syncing to iCloud?

Not from inside Harvestry today. The Save exports to iCloud Drive toggle in Settings → General only controls where exported study documents are saved. Library sync follows your iCloud sign-in. If iCloud isn't available on your Mac, everything is stored locally.

Compatibility

Which Macs are supported?

Any Apple Silicon Mac running macOS 15 or later. Harvestry uses the Apple Neural Engine for transcription, which requires an M-series chip. Intel Macs are not supported.

Does Harvestry work without an internet connection?

Yes, for local video files. Transcription, translation, screenshot capture, and export all run entirely offline. An internet connection is only needed to:

Troubleshooting

The app seemed to freeze briefly while transcribing a long lecture — is that a bug?

Not exactly. Harvestry transcribes entirely on your Mac using the Apple Neural Engine. For videos longer than about an hour, WhisperKit runs a large amount of parallel work in the background, and occasionally that causes a moment where the app appears unresponsive for a second or two before recovering on its own. Nothing is lost and transcription continues normally — the app is catching up with a backlog of background processing. For multi-hour lectures, leaving the app running in the background and checking back later is the most reliable approach.

A YouTube download failed with a 403 error — what do I do?

There are two common causes:

  1. You're connected to a VPN. YouTube frequently blocks download requests from VPN exit nodes. Disconnect from your VPN and try the download again — this is the most common cause.
  2. yt-dlp is out of date. YouTube periodically changes how it handles requests, and older yt-dlp builds start getting blocked as a result. Open Settings → Video Downloader and click Check for Updates to pull the latest release automatically.

If the error persists after both steps, try the download once more — YouTube occasionally returns transient 403s that clear on a second attempt.