OpenTypeless
Open-source dictation with choice of speech and language model, MIT licensed
Open-source dictation app, released under the MIT license, that lets users choose which speech recognition engine and language model it runs on. Available for Mac, Windows and Linux. Can run fully locally or connect to cloud models depending on configuration.
Index summary: OpenTypeless is listed in the The Dictation Index registry but is not ranked: 11 qualifying posts, 50 needed.
What people say
Example posts from the sample, each linked to the original on X and edged with the label it was given. A handful of posts cannot stand in for 11, so read the score and the topic table first. Quoted here and want it removed?
OpenTypeless works in whatever app you already have open. Press a shortcut, say the messy version, and it drops cleaned-up text into Slack, Gmail, or Cursor. Highlight a sentence and you can rewrite that by voice too.
Say goodbye to typing. OpenTypeless for macOS turns your voice into polished text anywhere: 🎙️ Hold shortcut & speak ✨ AI refines your thoughts 📥 Injects text right at your cursor 🔒 100% on-device or OpenAI/Anthropic LLM Try it:
Great stream. I’ve had similar experiences using AI to dig through manuals and find obscure documentation, the more context you give it, the better. One thing that’s really improved my productivity is voice transcription. I modified OpenTypeless to fit my workflow, then use
Registry facts on this page were checked against OpenTypeless's own site on 2026-09-16. See something wrong, or represent OpenTypeless? Request a correction or send a reply.