OpenRouter and Dome
OpenRouter offers hundreds of models, and only the approved ones reach your agents. One OpenRouter key reaches hundreds of models. Dome holds the key and exposes only the models you add as connections. A rule refuses anything not marked approved, and every call is metered against an agent.
Dome provides model brokeringand routing for OpenRouter.
One key, held in Dome
OpenRouter's key opens hundreds of models, so it stays with Dome. Agents hold their own Dome keys and never see it.
Approved models only
Each model is its own connection, tagged when it's vetted. A rule refuses every connection without the tag.
A fallback for any provider
Put OpenRouter behind a direct provider in a pool. When the first fails, the same model answers through OpenRouter.
Connect OpenRouterin three steps.
Add each vetted model as a connection, put OpenRouter behind Groq, and connect agents to the Gateway.
Add the connection
The provider id is openrouter. Dome uses https://openrouter.ai/api/v1 and OpenRouter's own model ids.
shelldome models add gpt-oss-openrouter \ --provider openrouter \ --model openai/gpt-oss-120b \ --api-key "$OPENROUTER_API_KEY" \ --attributes '{"approved":true}' \ --gateway prod-gatewayPool it behind Groq
Groq serves the same open-weight model and goes first, and OpenRouter answers when it can't.
shelldome models pool create gpt-oss-120b \ --failover-max all --gateway prod-gateway dome models pool member add gpt-oss-120b gpt-oss-groq --priority 0dome models pool member add gpt-oss-120b gpt-oss-openrouter --priority 1Point your agent at the Gateway
Code written for OpenRouter's API keeps working. Change the base URL to the Gateway and the key to the agent's.
pythonfrom openai import OpenAI client = OpenAI(base_url=f"{GATEWAY_URL}/v1", api_key=DOME_AGENT_KEY)client.chat.completions.create( model="gpt-oss-120b", messages=[{"role": "user", "content": "Summarize these release notes."}],)
Only approvedmodels.
Applied to an agent, this refuses any connection not tagged approved. Adding a model to OpenRouter's catalog changes nothing until you vet it.
cedar
forbid (principal, action == Dome::Action::"llm:invoke", resource is Dome::LLMModel)unless { resource has approved && resource.approved == true };Bringing it alltogether.
Connecting OpenRouter to registered agents, tools, and identity in Dome completes a governed agent application.
Acting for
Agent
Control point
GatewayEvery call decided and audited
RulesGuardsQuotas
Models
OpenRouterProvider
GroqProviderCommon questionsabout OpenRouter.
More of whatDome works with.
Model
Claude Fable
Fable 5.1 from Anthropic and Amazon Bedrock in one failover pool, open to one group and capped by quota.
MCP server
GitHub
The GitHub MCP server behind the Tool Gateway, with rules that decide each call on its owner and repository.
Identity
Okta
Okta tokens verified on every agent call, so rules and audit name the person each agent acted for.
Runtime
LangGraph
LangGraph agents with their model calls on the Model Broker and their MCP tools on the Tool Gateway.
Client
Claude Code
Claude Code on a Dome Gateway with per-developer sign-in, rules on every tool call, and audit by name.
Agent service
TinyFish
TinyFish's web agents behind the Tool Gateway, with rules that decide each run on the site it targets.
Talk with ourFDE team.
Or start free, and register your first agent in minutes. Connect it through a gateway, give it rules, and see every call it makes in one audit trail.
Talk with our FDE team
Our engineers design, build and run governed agents for the part of your business you want to change. You see what they deliver.
LLM providers for AI agents: one governed path to every model
Connect a provider once. Its key stays in Dome, and every agent call to it is authorized, metered and audited.