Everything on privacy, recording compliance, platform support, and how Xtify's local memory differs from cloud tools.
Every recording is saved to your Mac first, in full. The cloud is only used as a processing step for automatic speech recognition (ASR) — audio is sent up, transcribed, and the transcript comes back. We do not keep a copy of your audio or transcript on our servers afterward. Your local disk is the only permanent copy.
No. Xtify routes transcription and language-model extraction through enterprise-tier ASR and LLM services. Enterprise agreements contractually exclude customer data from model training — this is a meaningful difference from consumer-tier or free APIs, where usage data can often be used to improve the underlying model.
Recording-consent rules vary by jurisdiction — some regions only require one party's consent, others require everyone's. Xtify provides the recording and diarization tools; getting the appropriate consent from meeting participants where required is your responsibility. When in doubt, let participants know you're recording at the start of the call.
Nothing changes locally. Your recording and memory index already live on your device, not in our cloud. Any file we process is deleted from our servers right after — the only thing we keep is your account info, and that's erased from the cloud the moment you cancel, so nothing of yours is left behind.
All of them. Because Xtify captures system and microphone audio directly from your Mac rather than joining calls as a participant, it works the same way across Zoom, Google Meet, Microsoft Teams, Slack Huddle, WeCom, Feishu/Lark, Discord, phone calls, and any other meeting software — with no per-platform integration needed.
No. Xtify never sends a bot into your meeting. Recording happens locally from your Mac's audio, so there's no visible participant, no waiting-room approval, and no calendar invite required — useful for calls where a third-party bot wouldn't be welcome or possible.
Xtify currently runs on macOS 13 and later, on both Apple silicon and Intel Macs. An iOS version is in active development. There's no Windows version at this time.
Xtify identifies speakers by their voice, not by which microphone or device picked up the audio. Even two people talking into the same mic in a conference room come out as distinct, correctly labeled speakers in the transcript.
Most meeting-note tools store your transcripts and run their memory layer in the cloud, and expose it through a cloud-hosted MCP server that any authorized cloud agent can query. Xtify keeps recordings and its searchable memory index on your Mac, and exposes that memory through a local MCP server — so only agents running locally on your machine can read it. Cloud-based agents have no path in.
MCP (Model Context Protocol) is how AI agents securely query outside data sources. Xtify runs its MCP server locally, on your own Mac — so any AI agent you connect (like a local coding assistant or personal automation) can pull context from your meeting memory, without that memory ever being exposed to a third-party cloud service.
Individuals, freelancers, and consultants who sit in meetings across many different platforms and clients — not a team-wide software rollout that needs admin approval and per-seat licensing.
Yes — the Ultra plan partitions memory per client or workspace, so consultants running several engagements at once can recall context for one client without it bleeding into another.