General FAQ
Short answers to the questions people ask most before (and just after) they start using InkSpoke. Each answer links to a fuller page if you want the detail. If something here looks broken instead of just unclear, head to Troubleshooting.
About InkSpoke
What is InkSpoke?
InkSpoke is voice-first writing for every app you already use. You press a hotkey, speak, and InkSpoke transcribes your words on your own device, optionally polishes them with AI, and types the finished text straight into whatever app has your cursor — your editor, your email, a chat box, a terminal. No window-switching, no copy-paste.
The core loop looks like this on the desktop:
For the full picture, see What is InkSpoke?.
Which platforms does it run on?
| Platform | What you get |
|---|---|
| Desktop (Windows, macOS, Linux) | The flagship experience: push-to-talk dictation, AI refinement, Command Mode, meetings, voice control, and workspaces. |
| iOS | A dictation app plus the InkSpoke keyboard, which dictates into any app on your phone. |
| Android | The voice keyboard dictates into any app, with more features on the way (see below). |
| Web | Sign in to manage your account, API keys, active models, and your encrypted synced data. |
The desktop app is where every feature lands first. See How the pieces fit together for how they connect.
Do I need an account to use it?
No. InkSpoke works with no account at all — your device gets an anonymous identity and a free Pro trial the first time you run it, and dictation works entirely on your machine.
You'll want an account when you're ready to:
- Sync your workspaces and dictionaries across devices (opt-in, end-to-end encrypted).
- Buy Pro or Perpetual on the web.
- Manage API keys and active models from the web account.
You can sign in later from Account in the desktop app — it opens your browser to authenticate. See Signing in.
Privacy and offline use
Does InkSpoke work offline?
Yes — that's the default. On the desktop, speech recognition runs on a downloadable on-device model, and AI refinement can also run entirely on-device if you download a local model, so a full dictation can happen without any network at all. If you choose a cloud model, that step goes online instead.
On iOS and Android there is no on-device AI model for refinement — speech recognition can run on-device, but refinement uses InkSpoke's Platform (cloud) models or your own provider key.
Is my audio private? Does my voice leave my device?
By default your audio is transcribed on your own device and never uploaded. Whether anything leaves your machine depends entirely on which models you choose:
| Your choice | Where processing happens |
|---|---|
| On-device model | Fully local — audio and text stay on your device. |
| InkSpoke Platform model | Sent to InkSpoke's cloud for that step. |
| Bring-your-own-key (BYOK) | Sent to the provider whose key you configured (OpenAI-compatible). |
You pick these per step (speech and refinement independently). Read On-device vs. cloud and privacy for the full breakdown.
Where is my data stored, and how does sync work?
Your dictation history, recordings, and workspaces live locally unless you explicitly turn on cloud sync. Cloud Sync is off by default — you enable it under Account once signed in.
When you do sync, workspace content, dictionary terms, and settings are end-to-end encrypted before they leave your device, so InkSpoke's servers can't read them. On the web you can unlock and view — or delete — that encrypted data, but the decryption key never leaves your browser session. See Synced data and privacy and Account, sync, and updates.
Deleting your cloud synced data from the web doesn't touch the history on your desktop — that copy is separate and always yours.
Models and AI refinement
What models can I use?
InkSpoke groups models into three families, so you can trade off privacy, cost, and quality:
| Family | What it is | Notes |
|---|---|---|
| Platform | InkSpoke-hosted cloud models | The default for new accounts; nothing to download. |
| On-device | Downloadable Whisper (speech) and local LLM (refinement) models | Runs offline. On desktop, everything past the free speech model needs Pro. |
| BYOK | Your own OpenAI-compatible provider and API key | Requires Pro; your key is stored in your OS keychain. |
You set a default speech model and a default text (refinement) model, and workspaces can override them. See Models and providers and, on the web, Choosing your models.
Which model do I start with?
New users start on Whisper Small — a 244M on-device speech model that is free, offline, and requires no setup — paired with InkSpoke's Platform model for AI refinement. You can switch to larger on-device models or a cloud provider any time.
Do I need Pro to use AI refinement?
No — AI refinement is available on the Free plan using InkSpoke's Platform models, and there's a master toggle plus a quick hotkey (Alt + Shift + R, or ⌥ + Shift + R on macOS) to turn it on and off.
What Pro unlocks is the bigger tier of models: larger on-device Whisper and local LLM models, and bring-your-own-key providers. A few other features — Voice Commands and BYOK Command Mode among them — are also Pro-gated. See How refinement works.
Plans, trial, and billing
Is there a free plan?
Yes. Free gives you core dictation with InkSpoke's Platform speech and text models plus the on-device Whisper Small model, workspaces, and a monthly usage allowance. There are two paid tiers: Pro (a subscription) and Perpetual (a one-time purchase). Compare them on Plans and pricing, and check Usage and limits for what each allowance covers.
For current prices, the authoritative source is the pricing page at inkspoke.app/pricing. Enforced usage limits can differ from the headline plan copy, so treat the pricing page and your in-app usage view as the source of truth.
Is there a free trial, and how does it work?
Yes — the first time you use InkSpoke, your device gets a Pro trial automatically, with no account required. The trial is shaped around an audio allowance (roughly 20 minutes of dictation), and how the time limit applies depends on the platform:
| Where | Time limit | Audio allowance |
|---|---|---|
| Desktop | No time limit | ~20 minutes of dictation |
| Mobile | 14 days | ~20 minutes of dictation |
When the allowance (or, on mobile, the time) runs out, the trial ends and a desktop device drops back to Free. Full details are on The free trial.
How do I upgrade to Pro or buy Perpetual?
All purchases happen on the web through Stripe. From the desktop app, Account → Go Pro (or the upgrade prompt) opens your browser to check out. Perpetual is a one-time purchase: after buying, you activate a license key on each device from Account → Perpetual License. See Subscriptions and licenses and, on the web, Subscription and billing.
Can I buy or subscribe inside the app?
No. In-app purchase isn't available on iOS or Android — subscriptions and licenses are always bought on the web via Stripe. The mobile apps point you to the pricing page to check out, then your account (or an activated license) unlocks Pro on the device.
Can I buy extra credits or gift credits?
No. There are no credit packs to buy and no gift-credit flow. Your plan includes a usage allowance that refreshes each month; you don't top it up with one-off purchases. If you need more headroom, upgrade to Pro or Perpetual.
Mobile
Can I use InkSpoke on my phone?
Yes. The InkSpoke keyboard lets you dictate into any app on your phone. On iOS, because a keyboard extension can't record audio directly, tapping Start briefly launches the main app to capture your voice, then injects the transcription back at your cursor. iOS also pairs with your desktop over your local network (scan a QR code) so workspaces sync between them.
One iOS note: keyboard-to-workspace auto-matching is turned off, so you pick the workspace manually from the keyboard toolbar. See Set up InkSpoke on iOS, Dictate on your phone, and Syncing mobile with desktop.
What works on Android today?
Android is voice-keyboard-first, with more arriving over time. Here's the honest split:
| Working now | Coming soon |
|---|---|
| In-app dictation | Signing in to an account |
| Voice-keyboard dictation into any app | QR pairing / desktop sync |
| Workspaces, history, personal context and vocabulary | The typing keyboard and toolbar language/workspace pickers |
| On-device, cloud, and BYOK models | Command Mode / selected-text editing |
Account sign-in, desktop pairing, and the typing keyboard aren't wired up in the current Android build yet. Use the voice keyboard for dictation, and manage your account on the desktop or web for now.
See Set up InkSpoke on Android.
Getting help
How do I change my hotkeys?
All the global hotkeys are configurable under Settings → Configuration → Hotkeys on the desktop. The defaults are:
| Action | Windows / Linux | macOS |
|---|---|---|
| Start / stop dictation | Alt + Space | ⌥ + Space |
| Secondary activation | Ctrl + Shift + Space | ⌃ + ⇧ + Space |
| Command Mode | Alt + Shift + C | ⌥ + ⇧ + C |
| Toggle AI Refinement | Alt + Shift + R | ⌥ + ⇧ + R |
| Toggle Hands-Free | Alt + Shift + H | ⌥ + ⇧ + H |
The recorder warns you if a combo clashes with a reserved shortcut. See General and hotkeys.
How do I get support or send a diagnostic report?
The fastest way to give support useful context is the built-in diagnostic report. Open Account → Send Diagnostic Report on the desktop. It bundles system configuration and recent logs — no transcriptions, file contents, or personal data — and you can preview exactly what's included before sending. There's an optional voice self-test that reads back a fixed phrase so InkSpoke can check your mic and models end to end, and you get a reference ID when it's sent.
For self-service fixes first, start with Troubleshooting; for what the report contains, see History and diagnostics.
Next steps
- Troubleshooting — fixes for permissions, audio, models, and sync.
- Install and set up InkSpoke — get running in minutes.
- On-device vs. cloud and privacy — decide where your words are processed.
- Plans and pricing — compare Free, Pro, and Perpetual.