Frequently Asked Questions

Dikt is a Windows desktop speech-to-text application that combines local AI transcription with optional cloud services. It features a global hotkey for push-to-talk, automatic text injection at your cursor, and AI-powered cleanup of your transcriptions.

Dikt offers a 14-day free trial with all features included — cloud transcription and AI cleanup work out of the box, no API keys needed. After the trial, you can choose a BYOK license ($29 one-time) where you bring your own API keys, or a Pro subscription ($10/month) with everything included.

Not currently. Dikt is a Windows-only application that requires Windows 10 or later. Mac and Linux support may be considered in the future.

Yes. Local Whisper mode runs entirely on your machine with no internet connection required. Your audio never leaves your computer. Enable "Local Only" mode in Settings to disable all cloud features and work fully offline.

Dikt supports 100+ languages through OpenAI's Whisper models. You can set a specific language in Settings for best accuracy, or use auto-detection to let the model identify the language automatically.

AI Command Mode lets you transform selected text with voice instructions. Select text in any app, press Ctrl+Shift+Space, and speak an instruction like "make this more formal", "translate to Spanish", or "rewrite as bullet points". The AI replaces the selection with the result. It requires an OpenAI or Anthropic API key (BYOK) or a Pro subscription.

Cloud Sync is a Pro feature that syncs your custom vocabulary, snippets, context profiles, and AI prompt between the Dikt desktop app and your web dashboard at dikt.app/dashboard. Enable it in Settings > Advanced > Cloud Sync. Changes sync automatically using version-based conflict resolution.

Dikt has two overlay modes. The Notification Overlay shows transcription results as a brief semi-transparent popup that auto-dismisses. The Compact Overlay is a persistent always-on-top floating window showing recording status and audio levels. Both support configurable Type (Minimal/Line/Box), Style (Pill/Box/Text), Position (9 presets or custom drag), Opacity, Display (choose which monitor), and Size scaling. The compact overlay also features live waveform indicators (8 visual styles), Box-mode features (transcription preview, record/stop controls, mini history), click-to-open, lock position, auto-hide, and a letter-by-letter typing animation.

Voice profiles let multiple users share the same machine with separate settings, vocabulary, history, and statistics. Each profile stores its data in a separate directory. Switch profiles from the system tray > Profiles menu. The app restarts to load the new profile's data.

Dikt tracks consecutive days where you make at least one dictation. Your current streak, longest streak, and milestone badges (7-day, 14-day, 30-day, 90-day, 365-day) are shown on the Dashboard. It's a fun way to build a daily dictation habit.

Yes — Voice Profiles let each person have their own vocabulary list, history, and settings. Switch profiles from the tray icon or Settings → Voice Profiles. The app restarts to load the new profile's data.

Dikt tracks a daily dictation streak (like Duolingo). You earn milestone badges at 7, 14, 30, 90, and 365 consecutive days. View your streak on the Dashboard tab in Settings.

BYOK stands for 'Bring Your Own Keys.' You provide your own OpenAI and/or Anthropic API keys, which you get by creating accounts with those services. You pay those providers directly for API usage — Dikt doesn't add any markup.

Yes. API keys are encrypted using Windows DPAPI encryption. Local transcription never sends audio anywhere. When using cloud mode, audio is sent only to your chosen provider (OpenAI or Anthropic). Dikt does not collect, store, or share your transcription data.

Windows 10 or later (64-bit), .NET 8 Desktop Runtime (auto-installed if missing), approximately 200 MB of disk space, and a microphone. Internet is only required for cloud features.

Dikt offers three update channels: Stable (default, recommended), Beta (early access to upcoming features), and Canary (bleeding-edge, may be unstable). Change your channel in Settings > Advanced > Update Channel. Dikt auto-checks for updates and shows a notification bar when a new version is available.

The grace period lets you use push-to-talk mode without holding the hotkey. If you press and release the hotkey within the grace period (configurable in seconds), Dikt switches to toggle-like behavior — it keeps recording until you press the hotkey again. This is useful for longer dictations. Configure it in Settings > General. An optional audio cue plays when the grace period activates.

The notification overlay appears briefly after each transcription to show the result, then auto-dismisses. The compact overlay is a persistent floating window that stays on screen showing recording status and audio levels with live waveform visualization. Both overlays support configurable Type, Style, Position, Opacity, Display (multi-monitor), and Size. The compact overlay adds exclusive features: waveform indicators (8 styles including Mirror, Bars, Pulse, and more), Box-mode features (transcription preview, record/stop controls, mini history list), click-to-open, lock position, and auto-hide. Enable from the status bar toggle button or Settings.

If Dikt can't inject text because no window is focused or the target app blocks it, the transcription is queued instead of lost. The compact overlay shows a 'transcription queued' indicator — click it to inject into whatever window you focus next. The queue is cleared on app restart.

Wake word listening lets you say "Hey Dikt" to start recording hands-free — no hotkey needed. A lightweight background detector uses Windows built-in speech recognition to monitor your microphone and triggers recording when it hears the wake phrase. No model download required. Enable it in Settings > General > Wake Word.

The compact overlay can display a live audio waveform that responds to your voice during recording. There are 8 styles to choose from: Mirror (mirrored waveform above/below center), Bars (vertical bars), Pulse (smooth sine wave), Steps (staircase pattern), Dots (points at varying heights), Ribbon (filled envelope), Peaks (sharp zigzag spikes), or None (simple dot indicator). The waveform color changes by state: green when ready, red when recording, and blue when transcribing. Configure in Settings > General > Compact Overlay > Waveform Style.

A team workspace lets you share vocabulary, text snippets, and context profiles with your team members. When a team admin adds terms to the team vocabulary, every member gets them automatically via cloud sync. Personal settings always override team defaults when there's a conflict.

Team plans are billed per seat at $25/seat/month. When you create a team, you choose how many seats you need. You can add or remove seats anytime — billing adjusts automatically via Stripe.

Team owners and admins can invite members by email through the team management API or web dashboard. Invited members receive an email and can join the team once they have a Dikt account.

Team vocabulary is additive — your personal vocabulary is combined with team vocabulary, and both are used during transcription. For snippets and context profiles, team settings provide defaults, but your personal settings override on any key conflict. This means you always keep your personal customizations.

A Team Workspace lets your organization share vocabulary lists, text snippets, and voice profiles across all members. Individual users can still override any shared setting with their own preference.

Purchase a Team plan on the pricing page, then visit your account dashboard and click 'Create Team'. Invite members by email — each person uses their own Dikt license seat.

Team settings are the baseline. Any personal setting you configure in the app overrides the team default for you only. Shared vocabulary words are merged (team words + your personal words).

AI personas change the style and tone of your AI cleanup. Choose from 6 built-in personas: Technical Writer (precise, clear), Doctor (clinical notes), Journalist (engaging prose), Lawyer (formal legal language), Code Reviewer (preserves identifiers exactly), or Casual (light, conversational). You can also define your own persona with a custom system prompt. Switch personas in Settings > Speed & AI > Cleanup Persona.

Yes. AI Command Mode now supports multi-turn context — it remembers your last 3 instructions so you can refine your text iteratively. For example: 'make this more formal', then 'shorten it', then 'add bullet points'. Say 'start over' or 'clear history' to reset the context.

Word correction training lets you define wrong→correct pairs that are applied automatically on every future transcription. For example, if Whisper consistently transcribes a name as "Jon" when you mean "John", add a correction rule. Rules use case-insensitive whole-word matching and are applied as a post-processing pass before AI cleanup. Manage your corrections in Settings > Vocabulary > Word Corrections.

TTS (text-to-speech) preview reads your transcription aloud before injecting it — useful for proofreading long dictations. After transcription completes, Dikt reads the text back to you. Press your hotkey to confirm and inject, or press Escape to cancel. Choose between Windows built-in voices (free, offline) or ElevenLabs neural voices (high quality, requires API key). Configure in Settings > Speed & AI > TTS Preview.