Pick the right model
Create approved routes for different teams or tasks, then rank eligible models by priority, cost, or health. If a service goes down, use a backup you approved in advance.
Explore routingAI gateway and management reporting
Control AI spend. Show the return.
Give every team one safe way to use AI. Set who can use it, what they can spend, and what to do when a service fails. See the cost and the work it helped deliver across each team or company.
Try the Test Lab free. No card required; provider keys stay hidden after setup.
Portfolio operating reviewIllustrative management view · trailing 90 days
3 of 4 supportedAI operating cost$108.2K4 participating companies
Supported value$538KApproved by management
Cost coverage91%Mapped to owner and purpose
Cost against supported valueFour participating companies · trailing 90 days
Supported value AI cost
Atlas ServicesSupported
Northstar HealthReview
Harbor LogisticsSupported
Cedar ManufacturingIncomplete
Illustrative values show the intended decision hierarchy. Production reports identify source, owner, period, and support state for each measure.
OpenAIAnthropicAzure OpenAIGeminiBedrockCompatible APIs
One control point
Keep the tools. Change the operating rules.
Your apps use one web address. Change the models, spend limits, and backup plan in one place. No need to edit each app.
ApplicationsExisting tools and agents
AI Gateway HQIdentity · budget · policy
Approved modelSelected route or fallback
Control without disruption
Set the rules once. Keep work moving.
Teams keep the AI tools they know. You decide who can use them, how much they can spend, and which service takes each request.
Stop spending at the limit
Set a firm cap for the company, a team, or an app. The gateway checks each request before it can add another charge.
Control costApply company rules automatically
Start in watch mode. Turn on each rule when ready. Keep a clear record of each decision.
See governanceKeep AI keys in one safe place
Stop copying provider keys into apps. Give each app its own limited gateway key instead.
Review deploymentKeep the tools people already useChange providers in one place.
Change providers in one place.
Leave applications alone.
Codex, Claude Code, OpenCode, cURL, and compatible apps can connect to one web address. Change keys, limits, models, or backup routes without editing each app again.
Setup guides for Codex, Claude Code, OpenCode, and cURL OpenAI Responses and Anthropic Messages compatibility Self-service OIDC or SAML sign-in and SCIM lifecycle Company limits and approved fallback routes
See client setupimport os
from openai import OpenAI
client = OpenAI(
base_url="https://api.aigatewayhq.com/v1",
api_key=os.environ["AIGHQ_API_KEY"]
)
response = client.responses.create(
model="company-approved-fast",
input="Summarize the incident timeline"
)200 governedroute_cost-saverpolicy.allow
Built for shared ownership
Everyone sees the control they need.
Tech teams manage access and routes. Finance sees cost and approved results. Portfolio teams can compare the companies that take part. Each company keeps its keys, prompts, and financial records private.
Start in a test environment
Follow the safe-start path Observe first. Enforce after validation.
Add a limited test key. Review each rule and safety check before you send private or live work through the gateway.