For the complete documentation index, see llms.txt. This page is also available as Markdown.

Getting Started

SlinkyLayer can be used in three main ways:

  • Chat privately

  • Build private AI apps

  • Give agents private AI access

Each path uses the same core access model: choose an endpoint type, pay per request with USDC on Base through x402, and route the request through SlinkyLayer’s privacy layer.

1. Chat Privately

Users can access AI without creating an account, managing an API key, or attaching a billing profile on app.slinkylayer.ai

A typical private chat flow looks like this:

  • Choose endpoint type

  • Enter prompt

  • Approve x402 payment

  • Receive response

This is the simplest way to use SlinkyLayer.

2. Build Private AI Apps

Developers can use SlinkyLayer to add private AI access to applications without forcing users through traditional account and billing flows.

Instead of requiring users to sign up, create API keys, or subscribe to a plan, an application can send requests through SlinkyLayer and let each request pay for itself.

This is useful for:

  • Private AI chat apps

  • Crypto-native AI products

  • Research assistants

  • Developer tools

  • Agent dashboards

  • Privacy-first copilots

Developers can choose which endpoint types to support inside their applications:

  • ZDR endpoints for private requests

  • TEE endpoints for higher-assurance workflows

  • Standard endpoints for broader model access

SlinkyLayer is designed for autonomous agents that need to access AI without human-managed billing systems.

Agents can select an endpoint, make a request, authorize payment, and receive a response programmatically.

This makes SlinkyLayer useful for:

  • Autonomous workflows

  • Onchain agents

  • Machine-to-machine services

  • Agentic applications

  • Automated research systems

Agents do not need subscriptions, dashboards, static API keys, or manual billing operations.

They can access AI one request at a time.

Choosing an Endpoint Type

Before sending a request, choose the endpoint type that matches the use case.

  • ZDR endpoints For sensitive prompts where data retention should be minimized.

  • TEE endpoints For high-assurance workflows where runtime privacy matters.

  • Standard endpoints For general usage, experimentation, and broader model availability.

The endpoint type should be chosen based on the sensitivity of the request.

Paying for a Request

SlinkyLayer uses x402 payments with USDC on Base.

When a request requires payment, SlinkyLayer returns a payment requirement. Once payment is authorized and verified, the request continues.

In simple terms:

This removes the need for accounts, subscriptions, prepaid credits, or billing profiles.

Reserving Bandwidth

For occasional usage, pay-per-request access is enough.

For higher-volume usage, developers, users, and agents can stake $SLINKY to reserve bandwidth and unlock more predictable access to AI capacity.

Last updated