Grok and Dome

1. 01

   Connect xAI

   The provider id is xai. Dome holds the xAI key, and the research agent calls with its own.

   $ dome models add grok-xai \
     --provider xai \
     --model grok-4.7 \
     --api-key "$XAI_API_KEY" \
     --attributes '{"family":"grok"}' \
     --gateway prod-gateway

2. 02

   Connect OpenRouter

   OpenRouter names the model with an x-ai/ prefix.

   $ dome models add grok-openrouter \
     --provider openrouter \
     --model x-ai/grok-4.7 \
     --api-key "$OPENROUTER_API_KEY" \
     --attributes '{"family":"grok"}' \
     --gateway prod-gateway

3. 03

   Pool them for failover

   xAI at priority 0, OpenRouter at 1. The research agent sends grok-4.7 either way.

   $ dome models pool create grok-4.7 \
     --failover-max all --gateway prod-gateway
    
   $ dome models pool member add grok-4.7 grok-xai --priority 0
   $ dome models pool member add grok-4.7 grok-openrouter --priority 1

4. 04

   Point your agent at the Gateway

   xAI's API is OpenAI-compatible, so the OpenAI SDK needs only the Gateway URL and a Dome key.

   from openai import OpenAI
    
   client = OpenAI(base_url=f"{GATEWAY_URL}/v1", api_key=DOME_AGENT_KEY)
   client.chat.completions.create(
       model="grok-4.7",
       messages=[{"role": "user", "content": "Summarize this week's chip export news."}],
   )

5. 05

   Cap the spend

   One quota on the pool covers xAI and OpenRouter together.

   $ dome quotas set --subject pool --pool grok-4.7 \
     --unit provider_usd --limit 200 --window monthly

Commands and rules tested against a Dome workspace on October 1, 2026. For anything about Grok itself, see xAI's documentation.

forbid (principal, action == Dome::Action::"llm:invoke", resource is Dome::LLMModel)
when { resource has family && resource.family == "grok" }
unless { principal has act_as && principal.act_as.groups.contains("research") };