Can Jarvis Remember My Preferences Between Sessions?

Yes. Jarvis (getjarvis.eu) has persistent memory, so it remembers facts, preferences, and context across sessions rather than starting fresh each time. It is a desktop assistant for macOS 12+, Windows 10+, and Linux with a floating bar opened by Cmd+/ on macOS or Ctrl+/ on Windows. It can recall your preferred tone, your manager's name, recurring projects, and how you like emails formatted, so you do not repeat yourself. Memory is fully user-controlled: you can inspect, edit, or clear anything it has stored. Recall is applied as Jarvis routes work to frontier models from Anthropic, OpenAI, or Google. Your memory is never used to train any model, and connector tokens are encrypted with AES-256-GCM at rest under GDPR. A the free weekly 40 requests run out, then Pro is $16/month, Unlimited $32/month.

Persistent memory captures the kind of context an assistant needs to be genuinely useful over time: your standing preferences, the people and projects you reference often, decisions you have made, and patterns in how you work across apps. So when you say "email the usual team about this," Jarvis knows who that is, or when you draft in Slack, it already matches your tone. This is distinct from custom instructions, which are explicit rules you author; memory is learned context that accumulates from how you actually use it.

Crucially, memory is not a black box. You can open it, see what Jarvis has stored, correct anything that is wrong, and delete entries or wipe it entirely. That control matters for both accuracy and privacy, an assistant that remembers should also forget on command. If your role changes or a project ends, you prune the stale context so future drafts and actions reflect your current reality rather than yesterday's.

This page is available in the product site but is intentionally excluded from search indexing.

Capabilities