Aura is a high-performance voice-to-text refinement engine. Summon the capsule, speak naturally, and let Aura deliver a clean, structured draft directly into your active input field.
Aura removes filler words, fixes structure, and refines tone. It supports Chinese (Simplified) and English, with local or cloud model routing.
Turn messy spoken ideas into structured output with clear formatting.
Aura is minimal by design. The capsule stays out of your way while your text becomes clear, structured, and ready to ship.
One hotkey to start, one to finish. No UI noise.
Transforms spoken drafts into clean, professional writing.
Writes into your current input field or falls back to clipboard.
Run entirely offline or route through best-in-class providers.
Aura is optimized for people who think faster than they type. It turns spoken ideas into usable drafts without breaking your flow.
Lightweight, bottom-centered capsule mirrors premium voice tools.
English stays English, Chinese stays Simplified Chinese.
Use OpenAI, Anthropic, Gemini, DeepSeek, Qwen, GLM, Kimi, Minimax.
Settings dashboard stays clean and deliberately minimal.
“The capsule is frictionless. Aura feels like thought-to-text without the noise.”
Product lead, US startup“We shipped documentation faster because Aura cleaned speech into publishable prose.”
Engineering manager, EU teamAura is fully open-source and built with Tauri + Rust for performance and privacy. Ship it in your org, customize it, or keep it personal.
Clone the repo, configure your preferred models, and start speaking. Aura ships with a premium experience out of the box.