• Install agentgateway
    • LLM (OpenAI)
    • MCP servers
    • Non-agentic HTTP
    • Overview
    • Architecture
    • Set up an agentgateway
      • Listener overview
      • HTTP
      • HTTPS
      • mTLS (FrontendTLS)
      • TCP
      • Additional TLS settings
      • Customization options
      • Customize the gateway
      • Example configs
    • Helm
    • ArgoCD
    • Licensing
    • Install in air-gapped environments
      • Set up the UI
      • Explore the UI
      • Expose the UI via HTTPS
    • Sample app
    • Advanced settings
    • About
    • Model aliasing
      • Amazon Bedrock
      • Anthropic
      • Azure OpenAI
      • Gemini
      • OpenAI
      • OpenAI-compatible providers
      • Vertex AI
      • Multiple endpoints
      • Ollama
      • Mock LLM with httpbun
      • vLLM
    • API keys
    • Virtual keys
    • Load balancing
    • Model failover
    • Content-based routing
    • Streaming
    • Function calling
      • About guardrails
      • Regex filters
      • OpenAI moderation
      • AWS Bedrock Guardrails
      • Google Model Armor
        • Guardrail webhook guide
        • Webhook API reference
      • Multi-layered guardrails
    • Prompt enrichment
    • Prompt templates
    • Request transformations
    • Budget and spend limits
    • Rate limiting for LLMs
    • LLM cost tracking
    • CEL-based RBAC
    • Metrics and logs
    • About
    • Static MCP
    • Dynamic MCP
    • Virtual MCP
    • Connect via HTTPS
      • About MCP auth
      • Set up Keycloak
      • Set up MCP auth
      • About token exchange
        • Impersonation
        • Delegation
        • External IdP with Entra ID
        • About elicitations
        • Set up the elicitation infrastructure
        • Original elicitations
    • JWT auth for services
    • Control access to tools
    • Rate limiting for MCP
    • Stateful MCP
  • component_exchange Inference routing
    • About
    • Connect to an agent
    • Buffering
    • Direct responses
    • Dynamic Forward Proxy (DFP)
    • External processing (ExtProc)
    • gRPC routing
      • Request headers
      • Response headers
      • Early request header modification
      • Header
      • Host
      • HTTP method
      • Path
      • Query parameter
      • HTTPS redirect
      • Host redirect
      • Path redirects
      • Host rewrites
      • Path rewrites
      • Transformation templating language
      • Log CEL variables in access logs
      • Encode and decode headers
      • Filter request body fields
      • Inject LLM model headers
      • Forward request URL as header
      • Inject response body
      • Inject response headers
      • Rewrite path and method
      • Promote query parameters to headers
      • Remove request headers
      • Rewrite path segments
      • Update response status
      • Distributed tracing headers
      • Validate and default field values
    • Traffic splitting
    • HTTP connection settings
    • Keepalive
      • About
      • Request retries
      • Per-try timeout
      • About
      • Request timeouts
      • Idle timeouts
      • Per-try timeout
    • Access logging
    • BackendTLS
    • CORS
    • CSRF
      • API key auth
      • Basic auth
        • About OAuth
        • Access token validation
        • Authorization code
        • Keycloak as an IdP
        • MS Entra ID as an IdP
      • BYO ext auth service
      • About JWT auth
      • JWT auth for MCP services
      • Set up JWT auth
    • Token exchange
    • Local rate limiting
    • Global rate limiting
    • Debug your setup
    • Upgrade
    • Uninstall
    • OTel stack
    • Control plane metrics
    • Tracing
      • Solo Enterprise for agentgateway APIs
      • Kubernetes Gateway API
      • Agentgateway API
    • CEL expressions
      • Solo Enterprise for agentgateway CRDs
      • Solo Enterprise for agentgateway
      • Release notes
      • Changelog
    • Software Bill of Materials (SBOM)
    • Version support
  • open_in_new Migrate from Ingress
    • About Solo Support
    • Submit a request
    • Solo Enterprise for agentgateway
    • 2.3.x (latest)
    • 2.2.x
    • 2.1.x
    • GitHub
    • Twitter / X
  • to navigate
  • to select
  • to close
    • Home
    • Get started
    On this page

    Get started

    Install and run agentgateway, or route traffic to MCP, LLM, or HTTP backends.

    article

    Install agentgateway

    Install the agentgateway control plane in your Kubernetes cluster.

    article

    LLM (OpenAI)

    Route requests to OpenAI's chat completions API with agentgateway on Kubernetes.

    article

    MCP servers

    Connect to an MCP server and try tools with agentgateway on Kubernetes.

    article

    Non-agentic HTTP

    Route HTTP traffic to a backend such as httpbin with agentgateway on Kubernetes.

    Solo.io copyright 2026