About
English you can practise out loud, the moment you need it.
Most learners can read English. Many can write it. Far fewer can actually speak it without freezing — because the only real way to get fluent is to talk, and talking is the part real life never gives you enough of.
Why Tospeak exists
Booking a human tutor is a friction-heavy ritual: scheduling, payment up front, awkward intros, the social weight of disappointing them with your mistakes. The fastest way to ruin a speaking habit is to make every session feel like a performance.
Tospeak removes all of that. You open the page, click a tutor, and in three seconds you’re in a real video conversation with someone who will never be tired, never judge you, and never check their watch. You can quit after thirty seconds or talk for an hour. Both are fine.
What makes it different
- Real video tutors. Not a voice in a chat box, not a robot avatar. Faces and expressions that respond in real time.
- Speak any language; learn English. The tutors understand 30+ languages so beginners can ask questions in their own and get the answer in English.
- No booking, no signup, no payment. Try the product for as long as you want before committing anything.
- Privacy by default. No cookies, no tracking, no recording. Your camera and microphone stay on your device.
- Made for every age. Kids, teens, working adults, retirees. Same product, different doorways.
The tech, briefly
The tutor’s replies come from Anthropic’s Claude (currently Claude Haiku 4.5). The video is rendered by LiveAvatar. Speech-to-text uses your browser’s built-in Web Speech API, so audio of your voice is processed by Google rather than us. No transcripts are stored on our side.
Who’s behind this
Tospeak is an independent project. One person right now, building in public. If you have feedback, I read everything you send me at support@tospeak.ai.
No signup. Free to try.