Amical is the most direct Android competitor in this research — MIT-licensed, live on Google Play, with a genuinely useful context-aware formatting feature. One thing it doesn't disclose: whether its Android build transcribes on-device or in the cloud.
Amical (MIT, TypeScript/Electron, 1,468 ★) is genuinely on-device on macOS and Windows via Ollama, with cloud as an option there. On Android — where it's the most direct competitor to OpenWispr in this research — its architecture is different in one specific, unresolved way: neither amical.ai, its beta page, nor its quick-start docs state whether Android transcription happens on the device or in the cloud. Their marketing mentions cloud models "for increased accuracy and speed on mobile," which hints at cloud, but that's an inference, not a confirmed fact. This page won't guess, and it won't claim OpenWispr is "the only" on-device Android option because of that open question.
What is confirmed either way: both apps are MIT-licensed, both ship on Google Play, and Amical has a real differentiator — context-aware output formatting that detects the frontmost app and adapts tone (professional in Gmail, casual in Slack, clean in an IDE) across 100+ languages. iOS is invite-only TestFlight, not public.
This is the single most consequential unresolved question in the research behind this comparison. Amical's site, beta page, and quick-start docs never state where Android transcription runs. Their marketing copy about cloud models improving accuracy and speed "on mobile" is suggestive, not conclusive — it's consistent with cloud transcription, but it's also consistent with an optional cloud enhancement layered over a local base. Absent an APK teardown or a direct statement from Amical, calling it either way would be asserting something this research doesn't actually know.
This matters because it's the same trap this page is trying not to fall into on OpenWispr's behalf: the claim "the only free, MIT-licensed, genuinely on-device dictation app on Android" is a strong claim, and it would be false if said carelessly while Amical's status is unresolved. So it isn't made here.
Context-aware formatting — detecting the app you're dictating into and adjusting tone accordingly, across 100+ languages — is a shipped, genuinely useful feature OpenWispr doesn't have an equivalent for today. Amical also has an invite-only iOS TestFlight beta in motion; OpenWispr has no iOS build at all yet, in beta or otherwise.
Its two-tier hotkey pattern — Fn for push-to-talk, Fn+Space for hands-free continuous dictation — is also a well-designed piece of UX worth naming on its own merits, independent of the platform question above.
Open source and entirely on your device. Android from Google Play, macOS from GitHub — no account, no sign-up, just a download.