Config AI model, caption mode and voice settings
Current AI model
Backend default
Uses OPENROUTER_MODEL or openai/gpt-5.4-nano.
Caption mode
Manual / device dictation only
Use typing or your device keyboard dictation. KindVoya does not access microphone audio.
Server transcription model
OpenAI GPT-4o Mini Transcribe
Used only when Caption mode is Server AI transcription.
Speech output
Browser voice only
Uses this device/browser voice. No extra audio provider is called.
AI voice model
Sesame CSM 1B
Used only when Speech output is set to AI voice via backend.
Technical settings are managed separately. Keep this caption screen focused on communication.
Quick phrases Common care requests
Original Caption / Text
AI voice input
Manual typing or device keyboard dictation is active.
Checking translation usage limit...
Manual typing or your device keyboard dictation remains the lowest-risk workflow. Server AI transcription is v0.3-alpha: with consent, short audio is sent to KindVoya backend and OpenRouter STT, then the returned transcript is placed here for manual review before translation.
Latest Message
Original caption/text
No message translated yet.
Translation
Start a session and translate text to see output here.
Safety notes / preserved terms / confidence / risk level
Safety information will appear with each translation.
Session Transcript
This record was generated by an AI-assisted captioning and translation tool. It may contain transcription or translation errors. It is intended to support communication only and should not be treated as a verified clinical record unless reviewed and approved by an authorised healthcare or care professional. For diagnosis, treatment, medication changes, consent, safeguarding, mental capacity assessment, or urgent symptoms, use a qualified interpreter or clinician confirmation.
Export Integrity
No export signature has been created yet.