Show HN: Speechcap -Mac dictation with on-device transcription and cleanup

- Speechcap is a macOS app that provides real-time dictation entirely on-device, meaning no internet connection is needed and audio data never leaves the computer. This sets it apart from cloud-based dictation tools like those from Apple, Google, or OpenAI's Whisper API, which send audio to remote servers for processing. - "On-device transcription" implies the app uses local machine-learning models (likely Apple's native speech recognition or a lightweight third-party model) to convert speech to text in real time. - "Cleanup" refers to features that automatically remove filler words ("um," "uh"), repeated phrases, and false starts — a common pain point for writers, podcasters, and note-takers who find raw transcripts cluttered and hard to read. - The app appears to target power users, writers, developers, and accessibility-minded users who want privacy-first dictation without subscription fees or latency from cloud processing. - This was posted as a "Show HN" on Hacker News, a section of Y Combinator's news aggregator where makers present their projects for community feedback. It signals the product is early-stage and built by an independent developer.