Fireworks AI API and Dome
Open models on Fireworks AI, with every call governed. Fireworks AI serves Kimi, DeepSeek and gpt-oss under its own full model ids. Connect it to Dome and rules read who each agent is acting for, so a coding model stays with engineers. Together AI picks up Kimi K3 when Fireworks fails.
Dome provides model brokeringand routing for Fireworks AI.
No Fireworks key in agent code
Dome keeps the Fireworks key and adds it on the way out. Your agents carry Dome keys, one per agent.
Kimi K3 on two hosts
Together AI serves Kimi K3 too. With both in a pool, a Fireworks failure becomes a Together answer.
Models by team
Rules read the verified user an agent acts for. Keep a coding model for engineers, even when another agent asks for it.
Connect Fireworks AIin three steps.
Add Kimi K3 on Fireworks, pool it with Together AI, and send agents to the Gateway.
Add the connection
The provider id is fireworks. Dome uses https://api.fireworks.ai/inference/v1 and Fireworks' full model ids.
shelldome models add kimi-fireworks \ --provider fireworks \ --model accounts/fireworks/models/kimi-k3 \ --api-key "$FIREWORKS_API_KEY" \ --attributes '{"family":"kimi"}' \ --gateway prod-gatewayPool it with Together AI
Add moonshotai/Kimi-K3 on Together as a second connection, ranked behind Fireworks.
shelldome models pool create kimi-k3 \ --failover-max all --gateway prod-gateway dome models pool member add kimi-k3 kimi-fireworks --priority 0dome models pool member add kimi-k3 kimi-together --priority 1Point your agent at the Gateway
The OpenAI SDK works as is. Give it the Gateway URL and the agent's Dome key.
pythonfrom openai import OpenAI client = OpenAI(base_url=f"{GATEWAY_URL}/v1", api_key=DOME_AGENT_KEY)client.chat.completions.create( model="kimi-k3", messages=[{"role": "user", "content": "Review this diff for bugs."}],)
Kimifor engineering only.
Applied to an agent, this refuses connections tagged family: kimi unless the verified user is in the engineering group, and leaves other models untouched.
cedar
forbid (principal, action == Dome::Action::"llm:invoke", resource is Dome::LLMModel)when { resource has family && resource.family == "kimi" }unless { principal has act_as && principal.act_as.groups.contains("engineering") };Bringing it alltogether.
Connecting Fireworks AI 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
Common questionsabout Fireworks AI.
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.