Most AI speaking apps send your voice to a server and charge monthly rent. EdgeLingo runs the whole model on your iPhone instead — speech, conversation, and corrections, all on-device. Practice out loud in 9 languages, guided in your native language. Every turn gets a more natural rewrite, and reusable phrases return via spaced repetition. Describe a scene or snap a photo for a custom scenario. Works in airplane mode. One purchase — no account, no subscription, no per-message fees.
Framer AI AgentsDesign and publish professional sites with AI
Promoted
Maker
📌
Hey Product Hunt 👋
I'm a solo dev in Taiwan. I wanted low-stakes speaking practice I could do anywhere, but every app I tried uploaded my voice to a server and billed me monthly for the privilege. That felt backwards, so I spent this year getting the whole stack — speech recognition, the LLM, and TTS — running on-device.
The result: EdgeLingo works in airplane mode, never sends a byte of your audio anywhere, and costs one payment instead of a subscription. Nine languages, scenario-based conversations, natural-phrasing corrections, and spaced repetition on the phrases you actually used.
Trade-off worth naming: it needs a one-time model download and a recent iPhone. On-device isn't free.
Would love to hear what scenarios you'd want to rehearse — I'm reading every comment today.