Personal AI gets cross-model memory and 274 built-in actions your AIs can use for $39/month. Invite-only during beta — pricing applies after approval.
Solo users — 50M embed + 700K chat tokens/month, ~70 grounded conversations across Claude / ChatGPT / Gemini. Agent orchestration not included — upgrade to Pro for /agents access. Hit the cap? Opt into pay-as-you-go at $25/M chat / $0.35/M embed — you set the spend ceiling, no surprise bills.
Power users — 200M embed + 1.8M chat tokens/month + full agent orchestration (/agents). Daily Gmail / Drive / Slack / Notion sync, heavy cross-AI chat, AND ~30 multi-step agent runs typical (orchestrator + specialists share the chat budget — a typical run is ~50K tokens). Same pay-as-you-go overage as Personal.
Teams + regulated workloads — BYOK encryption, SSO/SAML, audit retention SLAs, multi-seat shared workspaces, custom token caps, unlimited agent runs (subject to negotiated cap), dedicated support. Contact sales for pricing.
Personal and Pro plans include a generous monthly AI allowance. If you exceed it, opt into pay-as-you-go and set your own monthly cap so there are no surprise bills.
Chat
$25/ 1M tokens
Charged only past your plan limit, when your assistants think and write.
Embedding
$0.35/ 1M tokens
Charged only past your plan limit, when new memory gets indexed.
Spend ceiling: you set the monthly cap in Settings → Billing. When you hit it, AI calls return a friendly “raise your cap” message instead of charging more.
No. You get a trial allotment to start, and we only ask for a card when you upgrade to a paid plan.
Yes. Cancel in one click from Settings → Billing. Your data stays searchable in read-only mode until the end of your billing period.
Claude (Desktop / Code / Web / Mobile), ChatGPT, Gemini (Web / CLI), Cursor, Codex, Kiro, Antigravity. The same memory layer and tool surface across every one.
Roughly 4 characters of English text. Plans give you orders of magnitude more embedding tokens (cheap) than chat tokens, since the typical workload is “index a lot, ask a few questions.”
Connect your AI in under a minute and see what a shared memory feels like.
Join the waitlistInvite-only during beta · pricing applies after approval