On-device · No subscription · Private
Your AI language partner lives inside your phone.
Real spoken conversations, natural-phrasing corrections, and the phrases that stick — on your device by default, with optional cloud AI that never hears your voice.
The line
Nothing crosses the line.
By default the model runs on your phone, so there is nothing to send and nowhere to send it — not your voice, not your photos, not a single sentence you practice.
Never hears your voice, sees your photos, or reads your chats.
Still never hears your voice or sees your photos. Only the text of your conversation is sent, answered, and not kept.
New in 2.0
A bigger model, when you want one.
Turn on cloud AI and Google Gemini takes over the conversation, corrections and hints. Your phone keeps the microphone, the voices and everything you've saved.
- Off until you turn it on
- Choose on the 2.0 welcome screen, or switch it in Settings any time.
- Text only, never kept
- Your speech becomes text on your phone; only that text goes through our servers to Gemini. We don't store it, and Google doesn't train on it.
- Lighter on your phone
- With cloud AI on, the on-device model isn't kept in memory.
- Offline still works
- No signal? EdgeLingo switches back to the on-device model by itself.
- Included with your purchase
- No subscription and no account. Each purchase gets a daily fair-use allowance; past it, practice simply continues on-device until the next day.
What you get
Everything you need to actually speak.
A calm, focused way to practice out loud — with a partner that adapts to you, on your device or, if you choose, in the cloud.
The loop
-
01
Speak
Real conversations
Talk your way through natural, scenario-based dialogues. Your partner keeps things flowing and stays in character.
-
02
Correct
Natural-phrasing fixes
After each turn, see a more natural way to say what you meant — gentle grammar and wording help, in context.
-
03
Retain
Phrases that stick
EdgeLingo pulls the reusable chunks from each chat and files them by topic — then builds a conversation whose goals are three of your own saved phrases, so they last.
And
- Any situation
- Describe a scene — or snap a photo — and get a practice scenario built for exactly what you want to rehearse.
- No connection required
- Once the models are on your phone, every feature works with no signal — and never costs a per-message fee.
- Pay once, no subscription
- One purchase, yours forever. No monthly plan, no per-message fees, no account — and cloud AI is included too, with a daily fair-use limit.
In action
One session, start to finish.
- 01Pick a scenario
- 02Hold to speak
- 03Read the natural phrasing
- 04Review, save the phrases worth keeping
About
Who makes EdgeLingo.
EdgeLingo is a one-person studio in Taiwan, founded in 2026. We build one thing: a language tutor that runs on your phone first — for adults who can already read a language but freeze the moment they have to speak it.
The idea came from a simple frustration — speaking practice needs to be low-stakes and always available, and every app we tried sent our voice to a server and charged rent for the privilege. So we put the model on the device instead: no account, no monthly fee, and your voice never leaves your phone. Version 2.0 adds an optional cloud mode for a bigger model — text only, and off until you turn it on.
I’m Eric Chen, a software engineer in Taiwan. From the idea to the build, EdgeLingo is so far entirely my own work. If you’d like to collaborate or need anything, do get in touch.
- Company
- EdgeLingo
- Founded
- 2026
- Founder
- Eric Chen (Jia Cing Chen)
- Based in
- New Taipei City, Taiwan
- Stage
- Live on the App Store
- Contact
- support@edgelingo.com