Sanity OS launches OTO AI, a one-time-purchase meeting assistant

The pitch behind OTO AI comes from a familiar frustration: leaving a meeting or interview and immediately thinking of the answer you should have given. Sanity OS says it tried existing solutions but found them disproportionately expensive and missing basics like adjustable text sizes, technical context, model choice, and the ability to distinguish real questions from conversational noise. OTO is its answer, built around listening to the conversation rather than recording it for later review.

The product differentiates itself from conventional meeting assistants in two ways. First, it works during the conversation, detecting questions and surfacing answers, explanations or key points without the user switching windows. Second, it lets users bring documents such as TXT, Markdown, DOCX and PDF files as background context, and with compatible multimodal models it can analyze shared screens, charts or code. Users connect their own API key, so the tool is not tied to a single AI ecosystem.

The commercial model is equally deliberate. Rather than a recurring subscription, OTO is sold once and owned for life, with users paying their AI provider directly for usage. Sanity OS frames this as a way to avoid another monthly software cost while still accessing leading models, though the actual cost will depend on how much API usage each user consumes.