It’s an AI-powered speaking practice application for learners aged 16+. It gives them a structured way to practise speaking, build confidence, and receive immediate feedback. 

Learners speak one‑to‑one with Calli, an avatar who adapts each conversation based on learner replies.
 

When Speaking companion is used for the first time, audio/microphone checks are completed on new devices or browsers. Learners are reminded that they are speaking with an avatar created by AI.

Activities include a goal, tips, and sometimes a short text or video. In some cases, learners can choose a focus area. They can end the conversation at any time, or the session will end when the time limit is reached.
 

After each activity, learners see a rewards summary, a CEFR vocabulary pie chart, a motivational message, and a transcript with audio playback. Learners can edit their transcript, and listen to their text read aloud. 
 

You can also preview Speaking Companion activities and try out the conversations. Because each activity adapts to the learner’s replies, your experience will be different to theirs. All conversations within the same activity will be different.

Speaking Companion data appears in class data views. It includes activity completion, number of attempts, and time spent talking. No scores are generated.
 

You can assign Speaking Companion activities using the usual assignment workflow. Assignments appear on teacher and learner pages and behave like assignments for other practice components.

For more information on assignments, read: How do I set homework assignments on Cambridge One?

Note: There is no lock-after-due-date option for Speaking Companion assignments.

You can restrict student’s access to Speaking Companion activities by creating locking/unlocking rules, like for other practice materials, if you wish.

For more information on managing student’s access, read: How do I create rules to manage student access to learning materials?