Diagnostic prerelease to help investigate Parakeet transcription failures and hangs.

What's New in VoiceInk 2.21

• 🦜 Added support for Parakeet Ultra model
• 🌐 Added support for OpenRouter transcription models
• 🤖 Added GPT-6 Luna and GPT-6 Sol enhancement models
• ⚙️ Support for custom AI enhancement model IDs
• ⌨️ Added double-tap recording shortcuts
• ↩️ Made Auto Send a global setting with automatic Return delivery
• 📋 Added mode duplication and context menu actions
• 🧠 Added local CLI support for Dictionary Auto Learn
• 📝 Improved dictionary editing and replacement rules
• 🛠️ Minor bug fixes and improvements

What's New in VoiceInk 2.20

• 🧠 Introduced the Dictionary Auto Learn feature
• 📦 Support for importing and exporting dictionaries
• 🕘 Faster access to recent transcriptions with Quick History
• 🖱️ Added support for mouse shortcuts
• 🗣️ Added support for Grok Voice Transcribe 2.0
• 🤖 Added support for Gemini 3.8 Flash and Qwen 3.8 27B enhancement models
• 🌍 Support for French localization
• ⚡ Improved OpenRouter performance and stability
• 🛠️ Minor bug fixes and improvements

New Features & Updates Video: https://youtu.be/29Wy0SkoWk8?si=0GGXKlNCRSOeYBhI

What's New in VoiceInk 2.13

• ✨ Added Gemini 3.5 Transcribe support
• ✨ Added SenseVoice Small support
• ⌨️ Improved cancellation handling
• 🎙️ Fixed the recorder panel issue — by @AlexSarbu
• 🎧 Fixed microphone routing issue in clamshell mode
• 🐛 Minor bug fixes and improvements

What's New in VoiceInk 2.11

• ✨ Added VoiceInk Refine for private, on-device transcription enhancement
• 🗣️ Added experimental Cohere Transcribe model support
• 🎙️ Fixed microphone issues when using a MacBook with the lid closed
• 🛠️ Minor improvements and bug fixes

• 🗣️ Fixed Apple Speech related issue
• ⚙️ Improved mode setup, availability, and model selection.
• 🌐 Added Dia browser support for URL-based mode triggers.
• 🤖 Added support for latest gemini enhancement models
• 🛠️ Minor improvements and bug fixes.

Introducing VoiceInk 2.0

A redesigned VoiceInk experience with powerful new workflows, broader model support, and improved reliability. This is public release version. I hope you'll love this new update.

• ✨ Introduced Modes for personalized workflows across different apps, websites, and tasks
• 🤖 Added new AI providers, custom models, and expanded transcription options
• 💬 Introduced VoiceInk Assistant for conversational follow-ups and contextual responses
• 🎤 Improved real-time transcription and recording reliability
• 🚀 Redesigned onboarding for faster and simpler setup
• 📊 Added productivity insights and history analysis
• 🌍 Added German and Simplified Chinese localization
• 🎨 Refreshed the interface with cleaner navigation and improved accessibility
• 🛠️ Added performance improvements, stability fixes, and general refinements

• 🚀 Support for Soniox v5 streaming model, Ink-2, and universal 3.5 pro model
• 📊 Redesigned dashboard and improved statistics view
• ⚙️ Trigger word support for modes
• 🐛 Minor bug fixes and performance improvements

• 🦜 Added new experimental transcription models (Nemotron multilingual & Parakeet Unified)
• 💻 New output mode to run shell scripts on transcription
• ✨ Bring back Trigger words for prompts
• 🌐 Added German localization
• ⚡ Soniox v5 async model support
• 🛠️ Fixed minor bugs & improvements

This is a beta release of VoiceInk 2.0 with a redesigned experience with a new Modes system, updated AI model/provider setup, and several workflow improvements.

Highlights

• New Modes system
• Updated AI model and provider setup
• Improved recorder and assistant workflows
• Minor improvements & bug fixes

Learn More: https://youtu.be/4TQfU_7uTIY