Hi guys.
I have been working on Hitoku Draft, an open-source, voice-first AI assistant that runs entirely locally. I posted about it already, and now it has also transcription with voice editing. Looking for feedback, as I found that outside tech circles other people still do not use this tech much.
It's context-aware, in the sense that it reads your screen, documents, and active app to understand what you're working on. You can ask about PDFs, reply to emails, create calendar events, use web search, editing text, all by voice.
You can download a compiled version for free with the code HITOKUHN2026 https://hitoku.me/draft/ (base price is 5 dollars)
It supports Gemma 4 and Qwen 3.5 for text generation, plus multiple STT backends (Parakeet, Qwen3-ASR).
Examples:
- Gemma4 in action, https://www.youtube.com/watch?v=OgfI-3YjEVU
- query a pdf document, https://www.youtube.com/watch?v=ggaDhut7FnU
- reply to email, https://www.youtube.com/watch?v=QFnHXMBp1gA
- and the usual voice dictation (with optional polishing)
I currently use it a lot with Claude Code and Logseq. Now with some friends we are also building a new cross-platform version. The goal is on the long run to have AI interactive local models serving people and professionals.
Active STT allows a "No STT loaded" option that mentions it requires a multimodal LLM like Gemma 4. Except even when I use Gemma 4 features, Ctrl+S to dictate doesn't work. Unless I Voice Edit then quickly Dictate as soon as it processes the silence. Sometimes if the Dictation is triggered on silence, it'll just choose to paste whatever text is on screen. There's no way to dismiss the popup with the text before it's ready to vanish on its own. There's no way to preview what the TTS voices sound like without triggering something to be said manually.
It seems like this will be a great tool soon, but currently there are very many rough edges that would benefit greatly from a nice heavy sanding pass.
reply