NetonAPIAn AI API relay designed from the operator's side

One key for every major model. Prices anchored to the official price, real cost tracked per channel, loss-making prices refused at save time, and the unit price shown is the unit price charged. Open source, self-hosted, a single native binary.

Quick start

Try Neton Cloud

GitHub

🏷️

Anchored to the official price

Sale price is "official × markup"; discounts for a group or a key account are "N% of official". When the official price moves, every price follows.

🛡️

Loss-making prices can't be saved

Each channel records its real cost. Every price is checked against every channel that would actually serve it before it is saved; a price health check lists the worst margin.

🧾

Shown price = charged price

Charge = tokens × the unit price the user sees. The pricing page, the console and the bill agree to the cent.

🔒

Two-phase ledger

Reserve, then settle, with the price snapshot frozen per request. Crashes recover automatically; unclear cases go to manual review — never auto-refunded, never double-charged.

🤝

Negotiated pricing

Find a user, set "N% of official" per model or in bulk, and see their price and your margin live. Dedicated lines and special rates attach to an API key.

🔁

Any protocol to any protocol

Native OpenAI, Anthropic and Gemini endpoints, converted in any direction with tool calls, images, thinking and cache usage preserved. Claude Code and official SDKs just work.

Neton Cloud: an API relay platform built on NetonAPI

Neton Cloud is an API relay operated with NetonAPI: sign up, create an API key and call GPT, Claude, Gemini, DeepSeek, Qwen, GLM, Kimi, Grok and more, billed at official prices.