Home

Gatewayz Wiki

Gatewayz is a universal AI inference gateway — one API that connects to 30+ AI provider gateways, giving access to 10,000+ models. Developers integrate once, get everything.

One API key, every AI model, automatic reliability, one bill.

---

New Here? Start with the Reading Path

These 6 docs tell the full story of Gatewayz — from "what is it" to "what still needs work." Read them in order. Each one builds on the last.

#	Document	What you'll learn	Time
1	Conceptual Model	What Gatewayz is, the architecture, all 10 system layers	15 min
2	Stability Definition	What "stable" means from user, product, and engineering perspectives	5 min
3	Conceptual Model Features	All 56 features the system must have (the spec)	20 min
4	Features	What's actually built today — 450+ endpoints across 17 systems	25 min
5	Delta Report	Spec vs reality: what's done, what's missing, what to fix (P0/P1/P2)	15 min
6	Features-Acceptance-Criteria	How we verify each feature is correctly implemented	20 min

After reading these 6 docs, you'll understand what Gatewayz is, what it should do, what it currently does, where the gaps are, and how we measure "done."

---

Testing & Verification

Once you understand the system, these docs cover how we test it.

Document	What it covers
Testing Guide	Start here — overview of all testing docs and how to run tests
Testing Plan	Manual API test cases (curl/Postman) for every feature
CM Unit Testing Plan	Unit tests derived from conceptual model claims
CM Unit Test Coverage Report	Which unit tests exist vs missing (196 dedicated tests, 5 xfail gaps)
Test Coverage Audit	Feature-by-feature automated test coverage

---

System Reference

Deep dives into individual systems. Read these when you need to understand or modify a specific area.

Security & Access

• API-Key-Management — Key encryption, scopes, IP allowlists, rotation
• Role-Based-Access-Control — Roles, permissions, hierarchy
• Activity-Logging — User action and system event logging
• Audit-System — Security audit trail, tamper protection

Billing & Monetization

• Stripe-Integration — Payments, webhooks, credit calculation
• Subscription-Plans — Plan tiers, lifecycle, credit allocation
• Free-Trial-System — Trial limits, lifecycle, conversion
• Coupon-System — Coupon creation, redemption, validation
• Referral-System — Referral codes, bonus rules, webhook flow

Monitoring & Observability

• Monitoring-System — Endpoints, Redis/Postgres data, circuit breakers
• Performance-Monitoring — TTFB, streaming duration, Prometheus histograms
• Prometheus-Setup — Metric categories, Railway deployment, PromQL
• Error-Monitoring — Sentry integration, alerts, performance spans

Features

• Chat-History — Persistent sessions, history injection
• Feature-Flags — Statsig gates, configs, experiments

Provider Integrations

• Alibaba-Integration — DashScope / Qwen model setup
• Google-Vertex-Setup — Gemini models, auth methods, regional endpoints

Operations

• Supabase-Migrations-CI — GitHub Actions CI/CD for migrations
• Testing-Workflows-Locally — Running GitHub Actions locally with act
• Troubleshooting — Common issues and resolutions

---

Data References

These are large generated documents — useful for lookup, not for reading start to finish.

• API-Mappings — Complete endpoint-to-code mapping (450+ endpoints)
• Test-Mapping — Test-to-feature mapping (~5,268 tests across 334 files)

---

Version 2.0.4 | Last updated: March 2026