Free & Open Source·Local processing, offline ready, data stays private

Speaking is the Most Natural Input

Just speak, leave the rest to AI.

Voconly converts your voice to text in real-time, understanding, polishing, and organizing it into ready-to-use content.

WindowsmacOS
Free & open source, unlimited useLocal & offline readyData never leaves your device4× Efficiency Boost
Voconly Demo
01 · Dictation

Voice dictation, unlimited and free.
As it should be.

No subscription, no crippled tier, no "3 left today". And on top of free, this is what daily use feels like.

There is no quota

Recognition runs on your own machine. No servers, no bills. No counts, no time limits, no sign-up — from the moment it installs, it is yours.

Used today47 / ∞
Remaining∞

Words as you speak

Text appears while you speak, ready the moment you stop. Recognition runs locally — no spinners, no cloud queue.

Filler words, gone

"Um", "uh", "I mean" — stripped before text reaches your cursor, punctuation handled. Want to keep your voice? Light polish only removes the fillers.

99+ languages, mix freely

Auto-detects 99+ languages and dialects. Switch between Chinese and English mid-sentence — no input-method juggling.

Audio never leaves

Meetings, ideas, trade secrets — processed locally, never uploaded. Open source, auditable.

Lands where your cursor is

Works anywhere you can type, without breaking your flow.

WordWeChatVS CodeNotionBrowserMailSlackFigma
02 · Key Workflows

One Key, One Workflow

Bind a different key to each scenario — press to speak, release to done. Every key is yours to define.

BoundClick a keycap
Right Ctrl

Professional Polish

Default

Turns loose speech into a rigorous, formal draft.

You say: This plan is okay I guess, it just feels a bit complicated.

The plan is viable overall but overly complex — streamline the core flow to reduce maintenance.

1Press the key2Speak into the mic3Release — text lands at cursor
Works out of the box: Right Alt for light polish, Right Ctrl for professional polish. More bindable keys: WinCaps LockF1–F14InsertPauseMenu
03 · Model Choice

Your Voice, Your Choice

Local or cloud, model or provider — you decide.

ASR

WhisperWhisper
SenseVoiceSenseVoice
ParakeetParakeet
Qwen-ASRQwen-ASR
CohereCohere
NemotronNemotron

LLM

OllamaOllama
OpenAIOpenAI
ClaudeClaude
GeminiGemini
DeepSeekDeepSeek
QwenQwen
GLMGLM
KimiKimi
GroqGroq

Privacy

Local processing
No account required
Works offline
Open source

Languages

99+
languages supported
中文English日本語한국어EspañolFrançaisDeutschItalianoPortuguêsРусскийالعربيةहिन्दीTiếng ViệtไทยNederlandsPolskiTürkçeBahasaSvenskaNorsk
04 · Pricing

Completely Free

Open source project, free forever

Open Source

Free & Open Source

$0

Local voice recognition, privacy safe

  • Unlimited voice inputs
  • Local processing, zero privacy risk
  • Smart polish & auto summary
  • Dual-platform: Windows/Mac
  • All future updates free
  • Open source, trusted & secure
More DownloadsView Source
Runs entirely locally, audio never uploadsWindows / Mac dual platformOpen source tech, trusted & secure
05 · FAQ

Got Questions?

If you can't find what you're looking for, get in touch.

View all questions

Save 30 minutes of typing every day

Local processing, offline ready, data safe

More Downloads

Windows · macOS dual platform