Choose translation direction
Open Actions → Translate → gear. Choose a target, or a pair that translates either language into the other. Text in a third language translates into the primary language. Search by language name or code.
Choose a translation provider
Automatic starts with Microsoft for fewer than 200 characters, Youdao for 200–3,999, and Apple Fast for 4,000 or more. Availability, supported direction and Allow online actions determine the bounded fallback. With online actions off, Automatic and Youdao use only Apple on-device translation.
Youdao supports 16 language/script options: Simplified Chinese, Traditional Chinese, English, Japanese, Korean, French, Spanish, Portuguese, Italian, German, Russian, Arabic, Thai, Dutch, Indonesian and Vietnamese. Unsupported languages or rejected directions use Microsoft or Apple when available. A temporary Youdao service error offers retry.
Apple Fast (on-device) and Apple Higher Quality (on-device) stay on your Mac. They use different Apple models and matching language downloads where supported by your macOS version and language pair; availability is checked in the app.
You can also select a configured AI provider. An explicitly chosen AI provider stays selected; reconnect it or change the choice if unavailable.
Install the matching Apple language packs
Open Actions → Translate → gear → Offline language packs. Select the Apple strategy you want, review the source → target status, and choose Install where offered. Two-language mode checks both directions. Changing the Apple strategy may require different downloads; installing one strategy does not install every model or language. Unsupported Apple pairs cannot be fixed by downloading a pack.
Review and use the translation
Show original text and Show Play button are independent options, off by default for new settings. Copy and Replace use only the translated output. Longer output scrolls after the automatic size limit; expand the popup for more room.
Choose speech voices
Enable Show Play button to see translation speech settings. Translation and OCR playback use installed macOS voices. Open Speech voices → Download Voices… for Apple voice downloads; these are separate from translation packs. Some Siri voices are not exposed to apps.
Say also offers online Edge voices, separately gated by Allow online actions. Spoken text goes directly to Microsoft. The consumer service can change or become unavailable; local speech is the fallback. Retry refreshes a failed voice list without deleting your saved voice.
Online actions and privacy
Microsoft, Youdao and online AI providers receive selected text only when their online translation runs with Allow online actions enabled. Apple translation stays on-device. Review language availability or read the privacy policy.
Translate a message and reply
Requires ActionClip 2.3.6. Translate a message first, then click Reply in the ordinary translation result. The popup shows Received, Your reply and Translated reply.
- The reply field receives focus automatically. Write in your preferred language.
- Use the language menus to choose the language you are writing and the language your recipient needs.
- Pause briefly while ActionClip translates your draft. If you continue typing, the newer draft supersedes the previous request.
- When ready, use the bottom Copy button to copy only the translated reply. Paste it into the original conversation yourself.
The popup grows with its content. Close it with the cross button; uncopied work may require confirmation. Replies use your existing translation provider and respect Allow online actions and installed Apple language availability. Online providers receive your draft for translation.
Follow each spoken word
In ActionClip 2.3.6, Say presents a read-only reading view. During playback, the whole current word receives a rounded highlight in your accent color. The view scrolls to keep the spoken word visible.
Use Play, Pause and Resume to control reading. Local speech uses macOS word boundaries; online Edge speech uses the timing supplied by Microsoft. Voice availability and timing depend on the selected service. Edge voices require Allow online actions and send the spoken text directly to Microsoft.