High-speed, push-to-talk voice typing with flawless Arabic & English code-switching. Hold your hotkey, speak your mind, and watch clean, formatted text paste into any active window.
Most AI dictation tools run on Chromium, hogging gigabytes of RAM and draining your laptop battery. FoxyFlow is compiled into lean machine code with near-zero system impact.
Zero background battery drain. Sits completely dormant until you hold your hotkey.
100x smaller than 500MB+ Electron apps. Installs in under 3 seconds without bloat.
Featherlight memory consumption. Your IDE, compiler, and games keep 100% of their speed.
Instant audio streaming to Gemini 2.5 Flash Audio or local Whisper for instant typing.
See how FoxyFlow compares directly to tools like Wispr Flow and typical Electron dictation apps.
| Capability / Feature | FoxyFlow | Wispr Flow & Electron Competitors |
|---|---|---|
| Account & Sign-up | ✓ Zero account needed. Just download & speak. | ✕ Mandatory cloud account, login & telemetry. |
| Pricing Model | ✓ 100% Free (BYOK: free tier or pennies/mo) | ✕ $12 – $24 / month recurring subscription |
| Binary Tech & Impact | ✓ Native Rust (~4MB binary, 0.1% CPU idle) | ✕ Heavy Electron/Chromium (500MB+, battery drain) |
| Offline Air-Gapped Mode | ✓ Embedded Faster-Whisper (Audio stays on PC) | ✕ Cloud-only; audio always sent across internet |
| API Key Security | ✓ Hardware Apple Keychain & Windows DPAPI | ✕ Stored on vendor's central servers |
| Arabic + English Switching | ✓ Native bilingual dialect & code-switching | ✕ Struggles with dialects and mid-sentence switching |
| Hotkey Ergonomics | ✓ Dual-Key Combos (Main + Custom Secondary Key, e.g. Ctrl+K) or Single Hold + Shortcut Guard + Shift Lock | ✕ Rigid, locked keyboard shortcuts; no custom key binding |
Hold your hotkey anywhere on your system to record. Release, and your speech is immediately formatted and pasted directly into your active window.
Switch between Arabic and English fluidly mid-sentence. Understands regional dialects (Levantine, Egyptian, Gulf, Darija) alongside technical terminology.
Your API keys are sealed directly inside macOS Apple Keychain or Windows DPAPI. Zero plain-text credentials stored on disk.
Use the ultra-fast Gemini 2.5 Flash Audio endpoint for state-of-the-art formatting, or switch to embedded local Whisper for air-gapped offline privacy.
Frameless bottom status pill that lives harmoniously at the edge of your screen. Unobtrusive visual feedback that never steals window focus.
Injects clean text anywhere: Slack, Notion, VS Code, Google Docs, Discord, WhatsApp Desktop, or native code editors without plugins.
Full keyboard control. Pair a Main Key (Ctrl / Alt / F8) with a Secondary Key (K, J, Space, or record ANY custom key), or choose single-key hold. Intelligent Shortcut Guard guarantees zero conflicts with system shortcuts (Ctrl+C/V/A).
Zero friction updating. FoxyFlow verifies release integrity with SHA-256 checksums, downloads updates silently in seconds, and restarts ready to dictate.
Dictate seamlessly in any app with intuitive hotkeys and ergonomic hands-free locking.
Activate dictation anywhere. On Windows, use a Dual-Key Combo (hold your Main Key and press your Secondary Key) or use Single-Key Hold (>300ms). The ambient HUD pill appears instantly at the bottom edge of your screen.
Need to speak at length? While holding your hotkey, tap Shift once to lock recording. Release all keys and talk freely without holding down any key.
Release your hotkey to stop recording. The AI engine cleans up fillers and pastes into your focused window. If you make a mistake or change your mind, hit Esc to instantly discard!
Fn (Globe) (>300ms) anywhere to dictate!
Control or Alt (Left or Right) (>300ms) anywhere to dictate!
No. FoxyFlow requires zero account registration, no email, and no credit card. It is completely free to download and use. You bring your own API key (e.g. Google Gemini’s generous free tier) or use the built-in offline engine. You never pay a recurring monthly SaaS fee.
FoxyFlow uses Gemini 2.5 Flash Audio API fine-tuned with smart multilingual prompt engineering. Unlike standard STT that forces a single language, it expects Arabic speakers to speak conversational dialects and English technical terms simultaneously, formatting punctuation and mixed script naturally.
Yes! In Settings, you can switch mode to Offline (Whisper). FoxyFlow downloads a compact local Faster-Whisper model (~148 MB). Audio is processed directly on your machine's CPU with zero network calls, making it 100% private and air-gapped.
Most modern desktop tools are built using Electron (bundling an entire Chromium web browser and Node.js runtime), which consumes 500MB+ RAM and 300MB+ disk space. FoxyFlow is written in 100% compiled native Rust. It uses native OS audio APIs (`cpal`), hardware keychains, and consumes only 0.1% CPU at idle and ~20MB RAM.
Your API keys are never stored in plain-text on your hard drive or uploaded to any third-party server. On macOS, they are sealed inside the Apple Keychain. On Windows, they are encrypted with the Windows Data Protection API (DPAPI).