Last updated July 2026
Wapo Assistant can convert your spoken words into text. When Voice Input is used, the temporary audio recording is sent through Wapo’s existing AI processing infrastructure to generate a transcription.
Wapo does not automatically send the transcription. You can review, edit, delete, or replace the text before choosing to send it.
Wapo does not store the raw audio recording. If you send the resulting transcription, that text is handled under the same Assistant and AI Memory rules as a typed message.
Wapo sends prompts to third-party AI models through the Lovable AI Gateway, which currently routes to Google’s Gemini family of models. Content sent for processing is subject to the provider’s terms during that call.
We do not send your email address, password, or unrelated content. Uploaded files are fetched from your private storage bucket at analysis time only.
Wapo does not use your content to train AI models. We rely on the provider’s enterprise-grade endpoints, which are not opted into training by default.
AI output can be wrong, biased, or incomplete. Verify anything the assistant tells you before acting on it. Never rely on Wapo for legal, medical, financial, or safety-critical decisions.
Destructive actions (completing a task, deleting data) always require your explicit confirmation. Document suggestions are proposed for you to accept or dismiss — they are never created automatically.
Concerns about AI use? Email hello@getwapo.com.