Tell One AI. Every AI Remembers.

Nikhil BhimaFounder2 min read
Side by side screenshots: ChatGPT being told 'I'm going vegan this month' on the left, and Claude recalling it unprompted while writing a fully plant-based meal plan on the right.

Tell ChatGPT you're going vegan this month. Then open Claude, a different product from a different company, and ask for a simple meal plan.

It comes back fully plant based. Claude's first words: "Great news, I have your context."

That's what the screenshots above show, captured on a brand-new account with nothing pre-loaded. One sentence to ChatGPT, then Claude planning the whole month around it. No copy-paste in between. The word vegan was never typed into Claude.

What's happening

Both AIs are connected to Tabula, one shared memory that belongs to you, not to either of them.

Mention something durable in ChatGPT and it saves the fact to Tabula. Ask Claude a question that needs it and Claude checks Tabula first. One AI writes, another reads. That's the whole product:

Tell one AI. Every AI remembers.

Why this matters

You already repeat yourself to your AIs. Your projects, your preferences, your constraints, how you like answers, explained again to every tool, forever.

Every AI company now ships its own memory, and every one of them keeps it inside its own walls. ChatGPT's memory stops working the moment you open Claude. Claude's doesn't follow you to Grok. The better each one gets at remembering you, the more locked in you are.

A shared memory flips that. Say things once. Switch models without starting over. Run three AIs side by side and have all of them know the same things about you.

Run the same test

Takes about two minutes per AI. Connect Tabula to ChatGPT, Claude, or Grok, tell one of them something true about you, then ask a different one a question that needs that fact.

Watching the second AI already know is the moment it clicks.

Everything your AIs save is yours: read it, edit it, delete it, export it, all from your dashboard.

Questions

How did Claude know what I told ChatGPT?
Both AIs are connected to Tabula, one shared memory. ChatGPT saved the fact there when it came up, and Claude checked there before answering. The two AIs never talk to each other; they read and write the same memory, which belongs to you.
Which AIs does this work with?
ChatGPT, Claude, Grok, Mistral, Perplexity, and Notion, plus anything else that supports connectors. The screenshots in this post are ChatGPT and Claude on a brand-new account.
Do I have to tell my AI to save things?
No. Once connected, your AI saves the durable facts you mention and recalls them when they matter. You can also say 'save this to Tabula' when you want something kept on purpose, and everything saved is visible, editable, and deletable on your dashboard.
Was this demo staged?
No. The screenshots come from a brand-new Tabula account connected to new ChatGPT and Claude accounts, captured in one take. Setup took about two minutes per AI, and you can run the same experiment yourself from the guide.

One memory that's yours, in every AI you use.