Windows 10/11

Speak code
into existence

The only voice tool that understands what you're building

Fast local transcription on GPU or CPU. AI reformatter can read your project context and write the prompt you meant. No cloud, no subscription.

Download free See pricing

Free for 7 days. Then $49 once — not $12/month. No credit card to start, no subscription ever.

> invoke
─ □ ✕
Why another dictation tool?

You think at 150 WPM.
You type at 40.

Cloud dictation tools add latency, cap your words, and send your audio to someone else's servers. Local tools exist but they're slow, clunky, and dumb—no context, no formatting, no auto-paste. And good luck getting a screenshot into Claude Code from your terminal.

Invoke runs Whisper on your GPU for sub-second transcription. Text appears at your cursor instantly. Turn on the AI reformatter and it reads your codebase to format clean prompts for Claude Code, Cursor, or whatever you use. All of it runs on your hardware.

Built by a developer who dictates 10,000+ prompts a month and was sick of copy-pasting raw Whisper output.

How it works
Three steps. One hotkey.
01

Hold & speak

Hold your hotkey and talk. Be messy. Say "um." Ramble. Start over mid-sentence. Invoke doesn't care.

→
02

Instant transcription

Whisper runs on your GPU. Sub-second. No cloud. No waiting. Your audio never leaves your machine.

→
03

Text appears

Auto-pasted at your cursor or copied to clipboard. Turn on the AI reformatter to get clean, context-aware prompts.

Features
What you get.
<1s

Sub-second on GPU

Whisper runs on your GPU — NVIDIA CUDA. A 30-second recording transcribes in under a second. CPU fallback included.

▶

Auto-paste

Text appears at your cursor the moment you stop speaking. No Ctrl+V needed. Just talk and it's there.

LOCAL

100% local & private

Your audio never leaves your machine. No cloud uploads, no accounts, no telemetry. Everything stays on disk.

∞

No word limits

Other tools cap you at 2,000 words/week. Invoke doesn't count your words. Dictate as much as you want.

{...}

AI prompt formatting

Optional: LLM reads your CLAUDE.md, .cursorrules, package.json, git log. Reformats speech into clean prompts for Claude Code, Cursor, Windsurf, Copilot.

🔒

Privacy mode

One click locks all processing to your machine. Use Ollama or LM Studio for local AI, or skip the LLM entirely. Works fully offline.

IMG

Screenshot to path

Snap a screenshot, press a hotkey — Invoke saves it and pastes the file path into your terminal. Image context for Claude Code without leaving your workflow.

8×

8 prompt modes

Output shaped for Claude Code, Cursor, Windsurf, Copilot, or Codex. Plus dedicated formats for git commits, PR descriptions, and raw cleanup.

aA

99+ languages & custom vocab

Whisper supports 99+ languages with auto-detect. Add custom vocabulary for your codebase — framework names, API endpoints, project terms — for sharper recognition.

Comparison
How Invoke stacks up.
Feature Invoke Wispr Flow SuperWhisper
Price $49 once / $79 lifetime updates $12/mo ($144/yr) $8.49/mo or $250 lifetime
Free trial 7 days, full features 2k words/wk 15 min
Speed <1 sec (GPU) 1-3 sec (cloud) ~1 sec (local)
Auto-paste yes yes no
Transcription Local (your GPU) Cloud Local
Privacy 100% local cloud upload local
AI reformatting LLM + project context Basic auto-edit —
Reads your project yes no no
Prompt modes 8 (Claude, Cursor, Codex, etc.) IDE extensions —
Screenshot paste file path — —
Platform Windows Win / Mac Mac-first, Win (limited)
Word limits none 2k/wk free none
Pricing
One price. No surprises.
MOST POPULAR
Invoke
$49
one-time — 1 year of updates
  • Sub-second GPU transcription
  • Auto-paste at cursor
  • Screenshot-to-path for Claude Code
  • Unlimited words, forever
  • 100% local & private
  • AI prompt reformatting (cloud + local)
  • Formatted for Claude Code, Cursor, Windsurf, Copilot, Codex + more
  • Context engine (CLAUDE.md, git log, etc.)
  • Custom vocabulary
  • Dictation history + search
Buy now

30-day money-back guarantee

or try free for 7 days

Invoke+
$79
one-time — lifetime updates
  • Everything in Invoke
  • Lifetime updates
  • Priority support
  • Early access to new features (IDE plugins, new modes)
  • Shape the roadmap
Buy now

30-day money-back guarantee

or try free for 7 days

7-day free trial, no credit card required. Pay once, use forever, no word limits.

Secure checkout via LemonSqueezy. 30-day money-back guarantee.

FAQ
Common questions.
Is there a free trial?
Yes. Download Invoke and use every feature free for 7 days. No credit card, no signup. After 7 days, enter a license key to keep using it.
Does it work without a GPU?
Yes. CPU mode is included and works out of the box—it's just slower (a few seconds instead of sub-second). A CUDA-capable GPU makes transcription near-instant.
What languages does it support?
All 99+ languages that Whisper supports. English is the default, but you can configure any language or enable auto-detection.
Does the AI reformatter send my code to the cloud?
Your audio always stays local. The reformatter calls an LLM API to clean up the transcript—you choose which provider (OpenRouter, Claude, etc.). For fully local operation, use Ollama or LM Studio, or turn on privacy mode to lock everything local with one click. No code files are ever uploaded.
How is this different from just running Whisper myself?
Running Whisper gives you a transcript. Invoke wraps that in a workflow — push-to-talk hotkey, auto-paste at your cursor, LLM reformatting that reads your project context, prompt modes for Claude Code / Cursor / Windsurf / Copilot / Codex, custom vocabulary, and dictation history. Whisper is the engine; Invoke is the car.
What are the system requirements?
For the Windows installer: Windows 10/11 (64-bit), 4 GB RAM, and a microphone. An NVIDIA GPU with CUDA 12.x is recommended but not required — CPU mode works out of the box. Internet needed for license activation and AI reformatter (unless you use Ollama or LM Studio). See full requirements.
What is screenshot-to-path?
Press a hotkey when you have an image on your clipboard (e.g., after Win+Shift+S). Invoke saves it as a PNG and pastes the file path into your terminal. This lets you share screenshots with Claude Code and other AI tools that accept image paths — no manual saving or typing file paths.
What platforms does Invoke support?
Invoke runs on Windows 10/11 (64-bit). An NVIDIA GPU is recommended for fastest transcription but not required — CPU mode works on any machine.
TRY IT NOW
Free for 7 days.

Download Invoke and start dictating in under a minute. No credit card, no signup.

Download free See pricing

Windows 10/11 · NVIDIA GPU recommended