gpt-oss and Dome
Run gpt-oss-120b from three providers under one name. gpt-oss is OpenAI's open-weight model, and that makes it the natural home for unattended work: nightly jobs, backfills, agents nobody is watching. Dome serves it from Groq, Cerebras and Together AI under one name, and keeps any call with no person behind it on open weights.
Dome provides model brokeringand routing for gpt-oss.
Three hosts, one name
Groq calls it openai/gpt-oss-120b and Cerebras calls it gpt-oss-120b. The pool gives agents one name to send.
Unattended jobs on open weights
Tag open-weight connections. When no verified person is behind a call, a rule refuses everything else.
Calls per caller
Give each agent and each person 2,000 calls a day on the Groq connection. One busy caller can't drain it for the rest.
gpt-oss with failoverin five steps.
Three hosts tagged weights: open, one pool, a rule for unattended calls, then a per-caller limit.
Connect Groq
On Groq the model is openai/gpt-oss-120b. Tag the connection weights: open; the key never leaves Dome.
shelldome models add gpt-oss-groq \ --provider groq \ --model openai/gpt-oss-120b \ --api-key "$GROQ_API_KEY" \ --attributes '{"weights":"open"}' \ --gateway prod-gatewayConnect Cerebras and Together AI
Each host serves the same weights under its own id, so tag them the same way.
shelldome models add gpt-oss-cerebras \ --provider cerebras \ --model gpt-oss-120b \ --api-key "$CEREBRAS_API_KEY" \ --attributes '{"weights":"open"}' \ --gateway prod-gateway dome models add gpt-oss-together \ --provider together \ --model openai/gpt-oss-120b \ --api-key "$TOGETHER_API_KEY" \ --attributes '{"weights":"open"}' \ --gateway prod-gatewayPool them for failover
Groq leads. A nightly job that loses Groq mid-run carries on through Cerebras, then Together.
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-cerebras --priority 1dome models pool member add gpt-oss-120b gpt-oss-together --priority 2Point your agent at the Gateway
The nightly job uses the OpenAI SDK with the Gateway as its base URL and its own Dome key.
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 last night's error logs."}],)Cap calls per caller
With --per-caller, a model quota applies to each agent and each verified person separately.
shelldome quotas set --subject model --model gpt-oss-groq --per-caller \ --unit calls --limit 2000 --window daily
Unattended jobsstay on open weights.
Applied to an agent with --agent, this refuses any connection not tagged weights: open when the call carries no acting-as identity. Calls made for a verified person are unaffected.
cedar
forbid (principal, action == Dome::Action::"llm:invoke", resource is Dome::LLMModel)when { !(principal has act_as) }unless { resource has weights && resource.weights == "open" };One call,two outcomes.
Switch the caller or the argument and watch the same call decide differently, with every decision landing in audit.
Model
agent nightly-batch · scheduled
llm:invoke(model: "gpt-oss-120b")
- Agentnightly-batch is registered and active
- CallerNo acting-as identity
- RuleConnection is tagged weights: open
DecisionAllowed
Bringing it alltogether.
Connecting gpt-oss to registered agents, tools, and identity in Dome completes a governed agent application.
Acting for
Control point
GatewayEvery call decided and audited
RulesGuardsQuotas
Models
Common questionsabout gpt-oss.
More of whatDome works with.
Provider
Anthropic
The Claude API behind the Model Broker: the key held in Dome, every call authorized, metered and audited.
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.
AI models for enterprise agents: failover, rules and quotas
Every model your agents call goes through the Model Broker. Pool providers for failover, decide who may call each model, and cap what it costs.