agentgateway review
A self-hosted Gateway API controller for securing and routing AI-native traffic.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
agentgateway is an open-source gateway proxy and Kubernetes Gateway API controller for teams operating AI-native infrastructure. It routes traffic between AI agents, LLM providers, tools, and services while providing authentication, authorization, TLS termination, rate limiting, traffic splitting, load balancing, and observability. Deployments can run through a standalone binary, Docker, or Kubernetes with the built-in control plane. Its self-hosted model makes it a fit for engineering teams that want direct control over gateway infrastructure rather than a managed service.
Its strongest distinction is the combination of conventional gateway functions with AI-specific connectivity. agentgateway can route LLM requests through an OpenAI-compatible API and connects with OpenAI, Anthropic, Gemini, and Amazon Bedrock. MCP and A2A gateway connectivity extends the scope to tools and agent-to-agent traffic, while OpenTelemetry metrics, logs, and tracing support operational visibility. JWT, API key, OAuth, external authentication, and authorization policies address access control, and health-aware Power of Two Choices load balancing adds workload-aware routing.
For Kubernetes traffic management, the published capabilities cover both Ingress and Gateway API models, along with TLS automation through cert-manager, weighted traffic splitting for A/B testing and canary deployments, and local or global HTTP rate limiting. These features give teams a broad control layer for self-hosted AI and service traffic. The tradeoff is operational responsibility: deployment, policy configuration, certificates, observability, and reliability remain part of the team's platform work. Choose agentgateway for an open-source, self-hosted gateway spanning Kubernetes, LLM providers, MCP, and A2A. Teams seeking a managed gateway experience or a narrower ingress-only controller should consider alternatives.
agentgateway pros and cons
- Where it wins
- Supports both Kubernetes Gateway API and standalone deployments
- Combines AI routing, MCP, A2A, security, and observability
- Includes TLS automation, canary routing, and rate limiting
- Where it doesn't
- Requires self-hosted deployment and operational ownership
- Focused on teams operating AI-native infrastructure
- Separate Solo Enterprise offering is not agentgateway pricing
agentgateway fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update