Version 1.0.0
Latest
First public release
SyncTrans 1.0.0 is the first version anyone outside the workshop can download. It turns a Mac into a live interpreter for a conversation happening in the room, and it lets you decide whether that conversation is processed on your own machine, in our cloud, or by an AI provider you already use.
New
- Three engine modes. On-device privacy runs recognition, translation and speech inside the app on models you download once, with no network needed afterwards. Worry-free cloud needs no downloads and no configuration. Custom service connects translation to OpenAI, Claude, Gemini or any compatible provider — including one you host yourself — with your API keys kept in the macOS Keychain.
- Dual-screen conversation mode. A second window carries another language to a second display, with each window remembering which physical screen it belongs to. Unplug that screen and the app falls back to the single-window layout you were using instead of stranding a window.
- Split-channel read-aloud. Assign each target language to the left ear, the right ear, or both, and two people can share one pair of headphones. The two ears are separate playback lanes, so both translations can be spoken at the same time rather than queueing.
- Three transcript layouts. Chat left and right, columns, and meeting notes — on ⌘1, ⌘2 and ⌘3, switchable mid-conversation.
- Projects and searchable history. Every turn is written to a project as it happens. Search runs across originals and translations, in one project or all of them, and each project remembers its own languages, engine mode and layout. Records stay on this Mac.
- Export as text, Markdown or subtitles. Plain text for pasting, parallel Markdown with speaker and time per turn, or an SRT subtitle track — all languages in one file, a single language, or one file per language.
- Glossary and rolling context. Lock a term to a fixed rendering in every enabled language, with an optional note. Choose how many completed turns each translation carries as context, from none up to a dozen.
- Fifteen conversation languages, seven at once. Arabic, Chinese, English, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Russian, Spanish, Thai and Vietnamese, with up to seven targets running together — each with its own voice.
- Two speakers on one microphone. Set a second language and completed turns are attributed to one speaker or the other by the language they were spoken in — no second microphone, no button to press before talking.
- Start without an account. A guest session gets you from opening the app to speaking with no form in between, on free starting credit. Registering adds its own trial grant.
- Seven interface languages. English, Chinese, Japanese, Spanish, French, Korean and Ukrainian, following whatever your Mac is set to, with an in-app override.
Privacy
- In on-device mode everything runs inside the app itself, and no request carrying your conversation is ever made.
- In cloud mode the service keeps usage records for billing — provider, duration, charge — and does not persist the source text, the translated text or the audio.
- Transcript history is written to your Mac in every mode and is never uploaded. Model downloads are encrypted in transit and integrity-checked file by file.
Requirements
- macOS 14 Sonoma or later, and a microphone.
- On-device mode needs an Apple Silicon Mac (M1 or newer) with roughly 16 GB of memory and 12 GB of free disk for the lightweight model set. The app measures your Mac and says which sets it can run before anything downloads.
- Dual-screen mode needs a genuine second display attached.
Download SyncTrans 1.0.0
Open the DMG and drag SyncTrans to your Applications folder.
Download .DMGmacOS 14+ · Apple Silicon · Version 1.0.0