Voice is macOS-only
Version 0.1.1 uses a native macOS speech sidecar. Windows reports dictation as unsupported. AI configuration cannot add speech capture to a platform where the release does not provide it.
AI and plan settings in release 0.1.1
The settings that are connected to the released voice-enhancement and entitlement flows.
Bring your own Gemini API key
Free users can choose the user-key mode, add and validate a Gemini API key, choose from the models offered by the app, and enable enhancement. The key is stored in the app's local settings and calls go to Google.
Review Google's terms, data handling, and charges before sending private material. Avoid dictating credentials, secrets, or sensitive source code.
Hosted enhancement
During the seven-day trial and with Pro, the app can send a captured transcript to the Ultragrip backend for Gemini-based enhancement. The backend manages the model, usage cap, and rate limits. This still requires macOS voice capture in the 0.1.1 desktop release.
Trial, Free, and Pro
Trial
Seven days from first install, with all profiles and hosted enhancement. No card is required to start.
Free
Claude Code profile and bring-your-own-key enhancement after the trial.
Pro, €4.99/month
All shipped profiles, mode switching, hosted enhancement, and priority email support.
Telemetry control
The released settings include an opt-out for usage telemetry. The documented payload covers app opens, controller connections, and dictation counts rather than transcript content. Read the current Privacy Policy for the complete public notice.
Controls not documented as working features
Version 0.1.1 displays deadzone, trigger-threshold, feedback, launch, and minimized-state controls, but the tagged frontend holds several of those values only in local component state. This documentation does not promise persistence or engine behavior that the release does not wire end to end.