Choosing Translation vs Speaking
Differences between AI Translation and AI Speaking, plus common questions.
Choose in one sentence
| Product | Who it suits | In one sentence |
|---|---|---|
| AI Translation | Simultaneous interpreting, following a meeting, listening to someone speak a foreign language | Listens to the other person, transcribes in real time and translates into a language you can read |
| AI Speaking | Practising a foreign language and delivery (including workplace delivery) | Listens to you, transcribes sentence by sentence and gives practice feedback |
Need to understand / translate what "someone else" says in a foreign language?
? AI Translation
Need to "speak it yourself" and get sentence-by-sentence feedback?
? AI Speaking
Comparing the two session areas
| AI Translation | AI Speaking | |
|---|---|---|
| Session 1 listens to | the other person | you |
| Session 1 main output | Source text + translation | Transcript + practice feedback |
| Session 2 main use | Text questions / multilingual output | Text practice dialogue |
| Session 2 read-aloud | ? | ? |
AI Translation
| Area | What you do | What the system does |
|---|---|---|
| Session 1 | Listen to the other person speaking a foreign language | Transcribes the source and translates it into your reading language |
| Session 2 | Ask questions / add context in text (optional) | Generates a reply/translation aimed at the other person; can be read aloud |
AI Speaking
| Area | What you do | What the system does |
|---|---|---|
| Session 1 | You speak into the microphone | Transcribes + gives practice feedback (not interpretation) |
| Session 2 | Practise in text (optional) | AI replies in text; can be read aloud |
The read-aloud button lives in session 2. Session 1 is transcription and feedback and does not play audio automatically.
Speaking Hub tabs: Self practice (default), Multi-party, and Expert session; see AI Speaking.
Common misconceptions
| Misconception | What actually happens |
|---|---|
| Speaking = translation plus one extra chat | Session 1 in Speaking is already practice feedback, not interpretation |
| Session 1 replies with voice automatically in both products | Read-aloud is mainly in session 2 |
| Session 1 is "interpretation" in both products | Only Translation listens to the other person and translates |
What to fill in under business settings
| Entry | Content |
|---|---|
| Settings ? YOYO assistant ? AI Language Partner | Auto-save interval, debug panel (client-side preferences) |
| Each product page ? business settings | Language direction, scene and vocabulary list, etc. (feeds the AI pipeline) |
Examples of scene and vocabulary list (optional):
- Translation:
Broad context: simultaneous interpreting in a business meeting/Vocabulary: agenda, quote, contract - Speaking:
Broad context: everyday Japanese conversation practice, N3/Topics: travel, ordering food, workplace topics
Common questions
Q: Why are there two session areas?
A: Session 1 handles the main voice flow (listening or speaking); session 2 handles text support (questions, practice dialogue).
Q: Why does only Translation show "source / translation"?
A: Only Translation listens to the other person. Speaking shows what you said + AI feedback.
Q: Where is read-aloud?
A: Next to the reply in session 2.
Q: How is this different from YOYO?
A: YOYO is a general assistant; the AI Language Partner is designed for long voice sessions, with chunking, an interpretation/feedback pipeline, bookings and metering.
How to use
AI Translation
- Open the translation page ? finish business settings
- Start a session or book a slot
- Pane 1 for live listen/translate; optional pane 2 for text follow-ups
AI Speaking (Hub)
- Open AI Speaking (default Self practice)
- Choose Talk live / Hold to talk / Type then send (Hold to talk and Type then send need no booking)
- Switch to Multi-party when you need role templates (colleague / client / friend)
- For deeper coaching, use Expert session (must book first)
See AI Speaking and Self practice.