← TellEddy

TellEddy Privacy

Where your voice and your text go — and where they don’t.

TellEddy listens to your microphone and reads text out loud. That’s inherently sensitive territory, so here’s exactly what happens to your data, broken out by feature.

The short version

• TellEddy has no servers. We don’t have an account system. We collect nothing.
• The default engines run on your Mac. Audio and text stay local.
• If you opt in to a cloud engine, your audio or text goes directly to that provider using your API key. Bear & Eddy never sees it.
• Listen Later items sync between your Mac and iPhone through your iCloud private database — encrypted by Apple, scoped to your Apple ID, invisible to us.
• Your API keys live in the macOS Keychain — encrypted by the OS, scoped to your account.
• No analytics, no telemetry, no crash uploads. If the app misbehaves you tell us; nothing phones home.

Dictation: turning your voice into text

Apple Speech (default)

Runs on-device on Apple Silicon Macs. Audio is processed by Apple’s SFSpeechRecognizer with the requiresOnDeviceRecognition flag set to true. Nothing leaves your machine — Apple isn’t involved past the framework, and Bear & Eddy isn’t involved at all.

WhisperKit (local)

Runs entirely on-device using OpenAI’s Whisper model, loaded into your Mac’s Neural Engine. The model is downloaded once from Hugging Face on first use; after that, transcription is local-only.

AI refinement (optional)

TellEddy can polish your dictated text using Anthropic’s Claude: punctuation cleanup, tone-aware rewrites, and notepad-mode polish that runs as you pause to think. This feature is opt-in and requires your own Anthropic API key.

When enabled, only the text you just dictated (and the requested tone) is sent to Anthropic using your key. TellEddy never proxies the request; we don’t see your text. Subject to Anthropic’s privacy policy.

Read-aloud: turning text into voice

When you pick a cloud engine, TellEddy opens a direct connection from your Mac to that provider. There’s no proxy, no relay, no Bear & Eddy server in the middle. We literally couldn’t see your text even if we wanted to.

Listen Later & the iPhone companion

Listen Later is the queue you build on your Mac and pick up on iPhone. The items in the queue — the URL, the title, the thumbnail, the rendered audio — sync through your iCloud private database (CloudKit).

Apple encrypts that database, scopes it to your Apple ID, and makes it inaccessible to anyone but you. Bear & Eddy doesn’t run sync infrastructure and has no read access to your queue. We have no servers in the loop and no way to see what you queued or listened to.

Sharing an item with a friend (via a /l/<id> link) is the one case where audio leaves your private database — only because you explicitly chose to share it. The shared link points to a small object you authorized; only people you give the link to can play it, and you can revoke the share at any time.

API keys (BYOK)

Cloud engines use your account and your key — we call this “bring-your-own-key.” You paste a key in once and TellEddy stores it in the macOS Keychain. The Keychain encrypts these at rest and only TellEddy can read them back on your machine.

Keys never get written to logs, never to disk in plaintext, never to crash reports (there aren’t any). If you uninstall TellEddy, the keys stay in your Keychain until you remove them yourself — they belong to you, not the app.

Recognition Hints & learn-from-dictation

TellEddy lets you add proper nouns and jargon as “Recognition Hints” so the dictation engine spells them right. Optionally, the app can learn frequently-used words from your dictation history and surface them as hint candidates.

When learn-from-dictation is enabled, TellEddy stores word frequency counts only — never full transcripts. A word has to appear several times before it’s surfaced. Stopwords are filtered. Everything lives in NSUserDefaults on your Mac. The feature is off by default; turn it on in Preferences if you want it.

If you grant the optional Contacts permission during onboarding, TellEddy reads names from your address book once and adds them to hints. Names stay on your Mac.

License codes

TellEddy is licensed with offline Ed25519-signed codes. We hand you a signed code; the app verifies the signature using a public key built into the binary. No server check, no phone-home — your license is just a signed string. The email address inside your license is shown back to you in the License window so you can confirm it’s yours; that email is not transmitted anywhere.

Permissions TellEddy asks for

• Microphone — to hear your dictation. Required.
• Speech Recognition — to use Apple Speech as the default engine. Required.
• Accessibility — to type text into other apps via synthesized keystrokes. Required for dictation to land at your cursor.
• Input Monitoring — to detect the hold-to-talk hotkey across the whole system.
• Contacts — optional. Names only, read once, stored in your Recognition Hints. You can skip this and the app works fine.

macOS grants these one at a time through its TCC permissions system. You can revoke any of them from System Settings → Privacy & Security at any time.

Updates

TellEddy checks for updates by fetching a small XML file from bearandeddy.com once a day. The request includes your app version and macOS version (standard browser User-Agent stuff) and nothing else. Updates themselves are EdDSA-signed; the app verifies the signature before applying them. You can turn auto-update checks off in Preferences.

What we don’t do

• No analytics framework. Not Google, not Mixpanel, not anything.
• No crash reporting. (We may add Sentry later; it’ll be opt-in and disclosed here first.)
• No account system. Nothing to sign up for, nothing to log out of.
• No user fingerprinting, no device IDs collected.
• No ads, no tracking pixels, no third-party SDKs that phone home.

Changes to this page

If we add a feature that changes any of the above — a new cloud engine, opt-in crash reporting, anything — we’ll update this page and call it out in the in-app release notes. The current version of TellEddy this page describes is the latest one on the TellEddy site.

Questions about privacy or security? Email drew@bearandeddy.com.