Gateways · Beta

Put a company spending boundary in front of OpenRouter.

Connect one OpenRouter API key. Give each app an approved model list, a hard spending limit, and its own usage record.

Create a free workspace Try the no-card Test Lab

How the connection works

How OpenRouter works with the gateway

Show only approved models

OpenRouter's catalog changes over time. Administrators choose which models each workload may use and publish stable aliases. Applications see the approved aliases, not the entire upstream catalog.

Stop each workload at its limit

Every app receives a scoped gateway key, budget, request rate, token rate, and concurrency limit. The gateway checks those limits before forwarding the request. OpenRouter's own credit and key limits still apply upstream.

Keep the two routing layers clear

AI Gateway HQ chooses whether a request goes to OpenRouter or another configured target. After OpenRouter receives it, OpenRouter may choose an underlying provider according to its own settings. Gateway evidence identifies OpenRouter as the selected outer target.

Current shared controls

What every supported connection includes

Encrypted BYOK credentials that remain hidden after setup Stable model aliases and compatible route targets Organization and key budgets, rate limits, and concurrency limits Observe, Shadow, and Enforce modes with metadata-only evidence

Test the exact OpenRouter model, provider preferences, request format, streaming behavior, and returned usage before production approval.

Client boundary

AIGHQ_BASE_URL=https://api.aigatewayhq.com/v1
AIGHQ_API_KEY=from-your-workspace
AIGHQ_MODEL=company-approved-fast

Applications use the gateway URL, a workload key, and a stable model alias. Administrators can change the approved OpenRouter model or move the alias to another compatible provider without distributing a new OpenRouter key.

Implemented boundary

Supported protocols

Select OpenRouter and enter its API key once. The connection test reads the current model catalog and does not generate tokens. Finish by assigning approved models to stable aliases and issuing a separate workload key to each app.

POST /v1/chat/completions POST /v1/responses POST /v1/embeddings OpenAI-compatible streaming GET /v1/models connection test

Compatibility follows the provider's current published interface. Review the provider documentation before approving OpenRouter for production.