Skip to main content

About PearPie

Privacy as architecture, not policy

Most AI platforms require you to create an account, store your conversations on their servers, and give you no say in where your data ends up. PearPie works differently.

Your conversations stay yours

Your conversations are not stored and not used for training. What you say to an AI model is between you and that model.

Private device sync

Your devices sync directly with each other over secure connections. No central server involved.

No account required

Use PearPie without signing up. A secure key on your device is your identity, no email, no password.

What PearPie is about

Privacy by design

No user identity is stored. No conversations are kept on any server. This isn't a policy, it's how the system is built.

Stronger together

Got friends or colleagues with a powerful GPU? Share models and resources with people you know over secure peer-to-peer connections. Your group gets more capable without depending on any central service.

Free to start, fair to scale

Running AI locally or sharing with friends is free, always. Credits are only needed for leading cloud models, and they're priced to be accessible.

Where it's headed

The goal: complete data sovereignty

The promises above hold today and they'll hold tomorrow. What's still evolving is how much of the infrastructure PearPie controls directly. Here's what that means and where it's going.

What's already true, and what's coming

Right now, your conversations are never stored, your devices sync privately, and you don't need an account. These things are built into the architecture and they're not going to change.

What's still improving is infrastructure sovereignty: the degree to which the hardware your data touches is fully under PearPie's control. Today, cloud model requests run on EU-based infrastructure within PearPie's own account. Your data stays European and ephemeral, but a third party still runs the GPUs.

That's the gap being closed. PearPie already supports running open-source models directly on your device, where your data never leaves your machine and you don't have to trust anyone. Over time, more models will run locally, a plugin ecosystem for developer tools is on the way, and the long-term goal is fully self-hosted GPU infrastructure running on European providers. At that point, even cloud processing runs on hardware PearPie operates.

The roadmap below tracks that journey, from "third-party involvement is minimised" to "there are no third parties."

Sovereignty roadmap

Phase 1
European-hosted cloud services
Cloud requests routed through EU-based providers within PearPie's own account. No conversation storage, no account required. Your data stays European and ephemeral.
EU processing No accounts No conversation storage Private device sync
Conversation privacy
100%
Device sync
100%
Infra sovereignty
35%
EU-hosted within PearPie's account, but third-party providers run the GPUs
Now
Local models on your own device
Run open-source models directly on your hardware. Full sovereignty, no technical knowledge required. For cloud models, the same EU-hosted promise continues.
On-device processing Peer-to-peer Zero network exposure
Conversation privacy
100%
Device sync
100%
Infra sovereignty
60%
100% for local models. Cloud models still via EU providers.
Phase 3
Expanded models and plugin ecosystem
More models available locally, including bring-your-own via Ollama. Plugin support for developer tools like Claude Code and OpenCode, all privacy-preserving by default. Credit usage is tracked for billing (linked to your public key, not an account), but your prompts, responses, and conversations are never stored.
More local models Plugin ecosystem Developer tools
Conversation privacy
100%
Device sync
100%
Infra sovereignty
75%
Most workflows can run locally. Cloud path improving.
Phase 4
Self-hosted European GPU infrastructure
PearPie-operated GPU clusters running on fully European providers. No third-party cloud in the data path. The hardware, the software, and the network, all controlled by PearPie, all European.
Self-hosted GPUs European providers Full stack control
Conversation privacy
100%
Device sync
100%
Infra sovereignty
100%
All data flows on PearPie-operated, European infrastructure
Vision
Complete data sovereignty
Every promise fulfilled across every model and every workflow. Local, peer-to-peer, or PearPie-hosted on European infrastructure. You choose how your AI runs. No third parties.
Complete control No third parties Your data, your rules
Conversation privacy
100%
Device sync
100%
Infra sovereignty
100%
Full sovereignty is the default, not the exception

Get in touch

Questions, ideas, or just want to say hello? Reach out at [email protected]