Default: model from server OPENAI_TRANSCRIPTION_MODEL. Kidney Care (tenant fields → kidneycare) forces whisper-1 + English translation. Chain with API 2 for notes.
Record directly from your microphone.
or click to browse
M4A · AMR · MP3 · WAV · FLAC · OGG
Only the audio file is required. Model and translation follow the rules below, starting with tenant domain name (tenant_domain_name).
Leave empty to use the server’s OPENAI_TRANSCRIPTION_MODEL (see live preview below). If this value normalizes to kidneycare, the API overrides to whisper-1 and English translation. No DynamoDB.
Form fields: user_id (email) and tenant_id (UUID or label). Used in server logs; tenant_id can also match kidneycare for routing if the domain field is not set.
ISO-639-1. Leave empty for auto-detect. Not used to pick the model.
Optional. Use with merge-chunk mode in Advanced.
Form field: tenant. Checked after tenant_domain_name and tenant_id. Same kidneycare rule.
Required for "Transcribe & Generate Notes".
Optional prompt context for the next segment.
When merge mode is on, send intermediate chunks with is_final_chunk off, then a final request with the last chunk and is_final_chunk on. Reuse the same Conversation ID.
Transcription results will appear here
Record audio or upload a file, then click a button on the left to get started