v1.4.0

Toro v1.4.0 Release Notes

Welcome to Toro v1.4.0! This major update brings a completely new invisible architecture to Toro, making the app much faster, significantly more secure, and completely frictionless to set up.

Tip

No More API Keys Needed!
You no longer need to manage, paste, or configure your own Deepgram or Gemini API keys. Toro now connects seamlessly through our brand new high-speed Cloudflare Edge Network. Just download and run!

What's New?

• Zero Configuration Setup: You can now start recording immediately out of the box. Toro handles all the complex AI routing and authentication in the background automatically.
• Invisible Edge Architecture: We moved the heavy lifting off your PC. Toro now connects to a globally distributed edge network, resulting in snappy AI formatting and lightning-fast dictation regardless of where you are in the world.
• Security First: Your device is now silently authenticated using a unique, secure device token that is safely kept in your operating system's native credential manager.
• Improved Background Reliability: We completely rewrote the core engine to be "asynchronous." For you, this simply means Toro will never freeze your computer and will reliably stop recording the precise millisecond you let go of the hotkey.
• Stable System Tray: Fixed a rare issue where the app could occasionally hang when trying to close it from the system tray.

How to Update

If you're already running Toro, just grab the latest toro.exe from the Releases page and replace your old one. All your settings are safely preserved!

v1.3.0

Toro v1.3.0 - The Cross-Platform Update 🌍

Toro v1.3.0 is a massive structural update that brings full native support to macOS and Linux, completely removing the terrible sudo requirements from earlier versions, and introducing an automated cloud build pipeline!

✨ What's New

🍏 Native macOS Support

• Removed sudo requirement: allowing Toro to leverage Apple's native Accessibility permission prompts instead of requiring root terminal access.
• Platform-Aware Pasting: Toro now intelligently detects your operating system and simulates the correct native pasting shortcut (Cmd+V for Mac, Ctrl+V for Windows).
• DMG Distribution: Added official .dmg packaging for simple drag-and-drop installation into your Applications folder.

📦 Installation

Windows Users:

1. Download Toro.exe from the assets below.
2. Double-click to run!

Mac Users:

1. Download Toro.dmg from the assets below.
2. Double-click the .dmg file and drag Toro.app into your Applications folder.
3. Note: On your first run, macOS will prompt you to grant Toro "Accessibility" permissions so it can listen for the Ctrl+Space hotkey.

v1.2.2

Toro v1.2.2 - Stability & Optimization Patch 🛠️

Toro v1.2.2 focuses heavily on refining the streaming architecture introduced in v1.2.1. We have squashed a few edge-case bugs that could cause audio chunks to leak between sessions and made key-release detection even faster!

✨ What's Fixed & Improved

⚡ Performance Tweaks

• Faster UI Teardown: Micro-optimized the key-release logic. The "Stop" sound now plays the exact millisecond you release the hotkey, completely bypassing hardware microphone teardown latency.

🧹 Chores & Cleanup

• Silenced Logs: Muted verbose connection pooling spam from httpx and httpcore to keep the app.log files clean and readable.
• UI Tweaks: Reordered the models in the system tray menu to place "Models" above "Quit Toro" for better UX.

📦 Installation

If you are upgrading from v1.2.1, you can safely replace your old .exe with the new one. Your existing .env file is fully compatible.

Happy dictating!

v1.2.1

Toro v1.2.1 - The Zero-Latency Streaming Update 🚀

We are incredibly excited to announce Toro v1.2.1! This update fundamentally changes the architecture of Toro, shifting from a post-recording processing model to a real-time, zero-latency streaming pipeline.

With this release, the total "AI lag" after releasing the hotkey has been reduced from ~5.0 seconds down to less than 1.0 second.

✨ What's New

⚡ Zero-Latency Streaming Architecture

• Real-Time: Audio is now transcribed in real-time while you speak . When you release the hotkey, the transcript is already complete!
• Background File Persistence:

🧠 Dynamic AI Model Switching

• System Tray Integration: You can now instantly switch between Gemini models ('Fast', 'Balanced', 'Accurate') directly from the Toro system tray menu. No app restart required!

v1.2.0

🚀 Toro v1.2.0 - The Observability Update
This release focuses on enterprise-grade monitoring, crash resilience, and performance tracing. We have integrated Sentry to ensure that Toro is faster, more stable, and easier to debug than ever before.

🛠️ New Features & Improvements

• Modernized Tracing: Migrated to the latest Sentry SDK v2.0+ standards using OpenTelemetry-compliant spans for high-precision profiling.
• Native Windows Crash Alerts: Implemented a native OS dialog box that triggers on fatal errors. If Toro crashes, it will now provide you with a unique Event ID to help the developer fix the issue.
• Cross-Platform User Identification: Added automatic user tagging. The app now identifies sessions by the OS username, allowing for personalized performance analysis.
• Zero-Config Telemetry: Added a fallback Sentry DSN so the application automatically reports health metrics out-of-the-box, while still allowing custom DSNs via .env.

v1.1.1

Initial public release