Spore

Spore is a distributed AI inference platform built by Spore Intel that lets individuals and developers run open-weight language models on hardware they already own, then reach those models securely from any device. Currently in open beta, it combines local inference (powered by llama.cpp) with an encrypted peer-to-peer network so users can both contribute and consume compute without relying on centralized cloud providers.

What It Is

Spore sits at the intersection of local AI inference and distributed compute sharing. Rather than forcing a choice between privacy (local tools) and convenience (cloud APIs), it provides an encrypted tunnel to your own nodes from anywhere, plus a community network for models your hardware can't run. The platform is OpenAI API-compatible, meaning existing code can point at Spore's endpoint with minimal changes.

How the Two Modes Work

Spore operates in two complementary modes:

• Personal mode — Run models on your own machine (macOS, Windows, or Linux) and access them from any device via end-to-end encryption. The server cannot read your data.
• Community mode — Tap the distributed network to run models that exceed your local hardware. Requests are encrypted and anonymized so processing nodes never learn who you are.

Users earn credits by letting their node serve community requests when idle, then spend those credits to access larger models on the network.

Setup Path

Getting started involves four steps the homepage describes: install the desktop app (one-click on macOS, Windows, or Linux), choose and host models on your hardware, optionally enable community serving to earn credits passively, and spend credits on the network when needed. On-device inference is powered by llama.cpp, the widely used open-source C/C++ LLM inference engine. The OpenAI-compatible API lets developers swap their base URL to https://api.sporeintel.com/api/v1 and keep existing code intact.

Privacy Architecture

Spore's privacy model is cryptographic rather than policy-based. Personal requests stay fully private; community requests are encrypted and anonymized end-to-end so that nodes processing them cannot identify the requester. The homepage explicitly states "the server can't read your data," distinguishing it from platforms that reserve the right to read prompts or train on them.

Why It Matters for Developers

The platform targets developers paying for cloud AI subscriptions while capable hardware sits idle at home. Key pain points it addresses include rate limits on hosted models, device-locked local tools like Ollama, and lack of genuine privacy on most platforms. The credit economy—earn by serving, spend to access—creates a self-sustaining incentive loop without requiring users to pay for compute they already own.

Current Status

Spore is currently in open beta. The web app is live at app.sporeintel.com and the desktop client is available for macOS, Windows, and Linux. The platform supports a curated set of open-weight models with no vendor lock-in, and business features are listed as "coming soon" on the features page.