Whisper Notes Blog: Latest Updates
Stay updated with the latest developments in offline voice transcription and AI speech recognition technology.
Recent Articles
May 12, 2026
SenseVoice: 52× Faster Chinese, Japanese & Korean Transcription on Mac
SenseVoice Small replaces Qwen3-ASR in Whisper Notes for Mac. GPU-accelerated via Apple MLX — 27 min Chinese podcast in 13.83s. Three-model benchmark comparison on M4 Pro.
March 19, 2026
Why Whisper Notes for Mac Left the App Store
Apple now rejects Mac App Store updates that use Accessibility permission. Whisper Notes moved to DMG to keep system-wide Fn voice typing. Free license for App Store purchasers.
March 7, 2026
Parakeet V3 Is Now the Default Mac Engine — 10x Faster Than Whisper
10x faster Mac dictation with NVIDIA Parakeet V3. 6.32% WER, zero hallucinations on silence. Now the default engine in Whisper Notes for Mac. Whisper models still available for 100+ languages.
December 3, 2025
Mac System-Wide Dictation with Whisper AI
Use Whisper AI for system-wide dictation on Mac. Hold Fn to type with your voice in any app—Gmail, Slack, VS Code. Offline, private, $6.99 lifetime.
August 2, 2025
Mistral Voxtral vs GPT-4o Audio – Open-Source Speech AI Benchmark
Discover Mistral's groundbreaking Voxtral models - the first open-source native multimodal speech recognition models with exceptional performance and competitive pricing that challenge industry standards.
May 29, 2025
Offline Speech to Text: Best Local AI Transcription Software
Three AI models run 100% on your Mac or iPhone — Parakeet V3, SenseVoice, and Whisper. No internet, no subscription. Compare speed, accuracy, and cost vs cloud transcription.
May 5, 2025
Offline Voice Memos for Privacy – Why I Switched from Cloud Apps
Discover why Whisper Notes is the top choice for secure, offline voice memos and meeting transcription, prioritizing privacy and accuracy with on-device Whisper AI.
March 15, 2025
Mac Whisper Notes – 10 Min Audio in 63 Seconds
Mac Whisper Notes rebuilt from scratch for macOS. Drag-and-drop audio import, keyboard shortcuts, timestamped results. M-series chips process 10 min audio in ~63 sec.
November 6, 2024
Whisper Large V3 Turbo vs V3: 5× Faster on Mac (Benchmark)
Whisper Large V3 Turbo vs V3 benchmark on Apple Silicon: 5× speed, 809M vs 1550M parameters, near-identical WER. Real numbers from M1/M2/M3 Macs.