Speak English with an AI tutor that helps you keep talking.

Responds to what you actually say

Tako is built around live voice conversation. It can follow everyday topics and work situations, ask a related question, and give you another chance to express the idea.

Lets you choose correction timing

Use Instant feedback when you want a supported correction while the sentence is still fresh. Use After-call feedback when protecting the flow matters more. Tako skips uncertain errors and ordinary spoken hesitation.

Adjusts how it speaks

Normal, Relaxed, and Slow modes change delivery through clearer phrasing and pauses. The aim is understandable speech, not mechanically stretching audio.

Looks across the speaking skills

• Grammar: patterns that change meaning or make the sentence harder to follow.
• Vocabulary: useful word choices and phrases connected to the topic.
• Pronunciation: audio-based observations when there is enough evidence.
• Fluency and clarity: how easily the listener can follow the idea.
• Interaction: answering, asking, and keeping a conversation moving.

What an AI tutor cannot promise

Speech recognition and feedback can be imperfect. Tako does not provide an accredited level, guarantee fluency, or replace a professional teacher. It gives you an accessible place to practise and reflect more often.

How voice data is handled

Tako does not store raw audio on its own server. Google processes live audio. Tako stores transcripts and grounded feedback so you can review conversations and see recent coaching observations.