On our gateway, every HTTP request payload is processed in volatile RAM and released when the response completes — no prompt database, vector store or training archive. Upstream model providers may apply their own logging; we disclose this transparently via audit headers on every response.
01 — Answers
Everything you need
to start building.
When you request
ethical-ai-auto, the gateway evaluates request intent in real time. Complex engineering, mathematics and agent tools route to capable reasoning models, while faster conversational requests use lower-cost endpoints.Yes. Set your base URL to
https://ethicalaico.com/v1 and pass your key as a Bearer token. Python, Node.js, LangChain, LlamaIndex, OpenClaw and AutoGen work with the compatible interface.The gateway is operated in Australia. We're verifying 100% renewable sources for upstream inference as our Phase 2 milestone — see the roadmap. Until verified, our audit headers disclose the actual power source on every response.
Use
ethical-ai-auto for everything — text, code, reasoning, tools and everyday agent work. One ID, flat pricing, no tiers.Token usage is metered per request and deducted from your account balance. There are no monthly subscriptions or minimum volume commitments—you pay for the compute consumed.
Yes. The gateway supports concurrent agent workloads, streaming responses and structured tool calls.
API keys use high-entropy values. You can manage, revoke or issue keys through the Account & Keys console.
Yes. For dedicated infrastructure, custom deployments or uptime commitments, email support@ethicalaico.com.