AI Gateway HQ vs. OpenRouter
Choose between broad hosted model access and a tighter company spending boundary.
- Pre-dispatch control
- Operator-ready evidence
- Prompt storage off by default
Choose AI Gateway HQ when each workload needs a funded ceiling and each company needs separate records. Choose OpenRouter for catalog breadth.
AI Gateway HQ checks the company, workload, policy, budget, route, and provider health before it sends a request. Strict mode reserves the budget first. Its BYOK price is based on successful requests, not a percentage of the model bill.
Compare what each gateway does.
Our column describes the working product. The OpenRouter column summarizes its linked documentation and pricing; it does not rely on anonymous review scores.
A compatible endpoint, encrypted BYOK pools, health-aware routing, shared cooldowns, and reason-coded fallback.
A hosted API whose current pricing page lists 500+ models and 80+ providers. It supports provider preferences, model fallbacks, automatic routing, prompt caching, and provider API keys.
Budget and capacity are reserved before forwarding. Settlement uses supported provider-reported usage, and promotional credit cannot fund server-paid models.
Guardrails can set daily, weekly, or monthly dollar limits for a member or API key. Workspace budgets can reset daily, weekly, monthly, or for a lifetime. OpenRouter's pricing page lists budgets and spend controls on paid plans, not its free plan.
OIDC, required MFA, least-privilege roles, workload keys, signed context, staged enforcement, and local attack screening.
Workspaces separate keys, routing, members, guardrails, and observability. Guardrails cover budgets, model and provider lists, data-retention policy, sensitive data, and prompt injection. The pricing page reserves SAML and managed policy enforcement for Enterprise.
AWS serverless deployment behind WAF, tenant-bound encryption, signed releases, metadata-only request records by default, tamper-evident exports, and approved support access.
OpenRouter runs the hosted service. It provides activity logs, exports, and management-key analytics. Its public material does not describe a customer-operated OpenRouter data plane.
A free BYOK proving tier, then $0.10 per 1,000 successful Flex requests. Inference bought through customer-owned accounts has no percentage markup.
Standard pay-as-you-go lists a 5.5% platform fee; Business lists 8%. Standard and Business each include a monthly BYOK allowance equal to $25,000 of list-price inference, followed by a 5% fee.
Consider OpenRouter when broad hosted model access is the main requirement. Its current pricing page lists more models and providers than AI Gateway HQ, and its routing layer can use OpenRouter credits or your provider API keys.
We reviewed OpenRouter's linked documentation and pricing on September 18, 2026. Vendors measure requests, logs, credits, infrastructure, and capacity differently, so compare the underlying units and recheck current terms before purchasing.