About us

ZGo Gateway

An open-source tool focused on AI API endpoint health checks.

Who we are

ZGo Gateway is a model endpoint gateway for AI developers. We believe that before wiring an API into your app you should confirm it “really works, how fast it responds, and whether its streaming cuts out midway”.

So we built this: paste a Base URL, auto-discover every model it offers, probe each with Chat and SSE (streaming), then produce a score, latency, availability and error classification as a shareable diagnostic report.

What we do

Connectivity probe
Discover /v1/models and verify Chat and streaming per model.
Gateway access
Call models through one endpoint with a ZGo-issued key; usage is billed from your account balance by token.
Diagnostic report
A–F scoring, latency distribution and error classes, produced as a shareable read-only report.

Our principles

We run both a model gateway and an endpoint checker. Requests through the ZGo gateway are proxied to upstream via our servers; checks run directly from your browser.

Zero-trust direct connection. By default your API key travels only between your browser and your own API service, never through our servers.

Minimal collection. Only server-side proxy checks route your key through our servers; if you have saved an endpoint configuration, it is stored encrypted, lookups return just the last four characters, and you can delete it at any time.

Open source

The project is fully open source — stars, issues and pull requests are welcome.

github.com/qingling-777/ZGo-API-doctor

Contact us

For any question, suggestion or partnership enquiry, reach us here:

support@zgoapi.cc