In development · macOS and iPhone
Dictation that keeps up when you switch 中文 and English mid-sentence.
Most dictation tools are built English-first. They handle a clean English sentence well, then fall apart the moment you say a Chinese clause with three English product names in it. Voicelna is built for that sentence.
- ~1.6 s
- Median time from the moment you stop speaking to cleaned text at your cursor. Raw text lands at ~0.5 s if you prefer speed over polish.
- One sentence
- No language switch, no mode toggle. Chinese and English in the same breath, with the English terms spelled the way you meant them.
- ≤ 40 words
- Your own vocabulary — names, models, jargon — sent with each dictation so the recogniser stops guessing. Stored on your device, not on a server.
Why bilingual dictation is the hard case.
It isn't about supporting two languages. It's about one sentence containing both.
The recogniser picks a lane
Most engines detect a language and commit to it. Say a Chinese sentence with "Kubernetes" and "staging" in it, and the English words come back as phonetic Chinese — or the whole line flips to English.
Proper nouns are guesswork
Product names, client names, internal jargon: the recogniser has never seen them. It substitutes the nearest common word, every single time, and you fix it by hand.
How Voicelna works.
Four steps, and only one of them is visible to you.
Hold a key and speak
Hold right Option in any app and talk. The connection opens the instant you press — the handshake hides inside your first words instead of landing on your face at the end.
Raw text comes back first
Streaming recognition, tuned for sentences that switch language mid-clause. Median 0.47 s from the moment you stop speaking.
A language model cleans it up
Punctuation, sentence breaks, filler words removed, lists actually formatted as lists. This is the step that turns speech into text you'd send.
It lands at your cursor
Not a transcript window you copy from. The finished text goes where you were already typing — email, editor, terminal, anywhere.
Where Voicelna sits.
| Built-in macOS dictation | English-first AI dictation | Voicelna | |
|---|---|---|---|
| Chinese-English in one sentence | Partial | Weak | Designed for it |
| Cleans up filler and punctuation | No | Yes | Yes |
| Learns your proper nouns | No | Rarely | Yes, ≤40 terms per dictation |
| Text lands at the cursor | Yes | Yes | Yes |
| Vocabulary stored on device | — | Usually server-side | On device |
Where your words go.
Audio goes to Aivmelna's own inference cluster, not to a third-party speech API. Your personal vocabulary — the names and terms Voicelna learns — stays in a file on your own machine; the server is sent at most the handful of terms relevant to the current dictation and stores none of them. Logs record counts, not content.
Questions.
Does Voicelna handle Chinese and English in the same sentence?
Yes — that is the case it is built for. You do not switch languages or modes. You speak one sentence containing both, and the English terms come back as English words rather than as phonetic Chinese.
How fast is it?
Median 1.6 seconds from the moment you stop speaking to cleaned text at your cursor, with a 90th percentile of 2.5 seconds. If you would rather have speed than polish, raw text lands at about 0.5 seconds and you can skip the clean-up step.
What is the difference between dictation and what Voicelna does?
Plain dictation gives you a transcript: no punctuation, no paragraph breaks, every "um" preserved. Voicelna runs that transcript through a language model first, so what reaches your cursor is text you could send without editing.
Can it learn names it keeps getting wrong?
Yes. Voicelna keeps a personal vocabulary of the terms you actually use, together with the wrong spellings the recogniser has produced for them. Both are sent with the next dictation, so a name you corrected once tends to stay correct. The list is capped on purpose — an oversized glossary makes accuracy worse, not better.
Which platforms does it run on?
macOS, with a global hotkey that works in any application, and iPhone, as a third-party keyboard so dictation works inside other apps.
Is it available yet?
Not yet. Voicelna is in development and being used daily by its author. Email hello@aivmelna.com to be told when it ships.