You're typing, not listening
Every line you write is a sentence you didn't hear. You walk out with a page of fragments and no real memory of what was decided.
Hit record. Voxxli transcribes every word as it's spoken, labels who said what, and hands you the summary, the decisions and the action items, without your audio ever leaving your iPhone.
The note-taking tax
You can pay attention or you can take notes. For years those were the only two options.
Every line you write is a sentence you didn't hear. You walk out with a page of fragments and no real memory of what was decided.
Forty-seven minutes of audio with no timestamps and no search. Scrubbing through it costs more time than the meeting did.
A wall of undifferentiated text is a transcript you have to relive to understand. Without speaker labels it's barely better than the raw audio.
Why Voxxli is different
Most meeting recorders ship your raw audio to a server to transcribe it. Voxxli does that work on Apple's on-device speech engine, in your hand. Here is where every part of a recording actually lives.
How it works
One tap and you're capturing, with a live waveform, live timer and live timestamps. Already have the audio? Import any audio or video file instead.
Words appear as they're spoken, in any of 50+ languages, with timestamps you can tap and speaker labels that sort out who said what.
Summaries, decisions and action items are waiting when you stop. Then ask it questions, or turn the whole thing into flashcards.
Features
On-device speech recognition turns speech into searchable, tappable text while the conversation is still happening.
Voxxli works out who's talking and labels each line, so you can see who committed to what.
"Let's ship onboarding first."
"Agreed, target Friday."
"I'll own the rollout."
Key takeaways, decisions and action items, extracted the moment you stop.
Any recording becomes a Q&A deck. Built for lectures and exam weeks.
Tap to flip
Ask a recording anything instead of scrubbing through it.
The same line, transcribed in whichever language the room is speaking. Import audio and video from elsewhere too, and export anything as PDF or Markdown.
Built for
Where it earns its place
Every one of these is a moment where the notes never happened. That's the gap Voxxli fills.
Questions
Yes. Recording and transcription run entirely on your device using Apple's built-in speech engine, and your recordings live in the app's own sandboxed storage. They are never uploaded to us or to anyone else.
A few of the AI extras (summaries, flashcards, chat and speaker labels) do need a connection, because that step is processed off your device from the transcript text. Your audio is never part of it. The privacy policy sets out exactly what is sent and when.
Recording and transcription do. Both run on-device, so they work in airplane mode, in a basement lecture hall, or anywhere with no signal. The AI extras (summaries, flashcards, chat and speaker labels) need a connection. Record offline now, generate the summary when you're back online.
Voxxli is free to download and free to start using. There's an optional Voxxli Pro subscription if you want unlimited recording, imports, summaries, flashcards and chat, but you don't need it to try the app, and there's no account to create either way.
Over 50, covering English, Spanish, French, German, Italian, Portuguese, Dutch, Polish, Romanian, Swedish, Japanese, Korean, Chinese, Arabic, Hindi, Turkish and many more. Available languages depend on the speech models your device supports.
Yes. Import existing audio or video files and Voxxli will transcribe, label and summarise them exactly as it would a live recording.
Export any transcript, summary or set of minutes as PDF or Markdown, and share it wherever you like.
Speaker identification works from the transcript text rather than from voice fingerprinting, so it infers turns from how the conversation reads. It's good at clean back-and-forth dialogue and less reliable with overlapping speech, long monologues, or recordings where the turns aren't obvious. Treat labels as a strong first pass you can correct, not as forensic attribution.
Not currently. Voxxli is built on Apple's on-device speech frameworks, so it runs on iPhone.
Get started
Free to download. No account to create. Your audio never leaves your phone.