Connection guide
How to connect Claude Code to your OpenAI costs
Connect Claude Code to your OpenAI costs through Costory's FinOps MCP. Claude Code gets OpenAI spend allocated to teams and features, next to the deploys and metrics that explain it.
What you can ask Claude Code about OpenAI
“What is our OpenAI spend by project and model this month?”
“Which team's API key drove last week's OpenAI spike?”
“Show total AI spend across OpenAI, Anthropic and Bedrock.”
“Which team owns most of this month's increase?”
“Set up a weekly Slack report of this spend for #finops.”
What Claude Code sees
Costory ingests OpenAI cost through OpenAI organization costs with a read-only Admin API key, normalizes it into one cost model alongside your other providers, and allocates it with projects, API keys, models. Through the MCP server, Claude Code can query that allocated cost, compare periods, look up deploys and other events in the same window, and create alerts, reports and dashboards in Costory. It has no write access to your cloud accounts.
Good to know for this setup
If Claude Code reports that the server “does not support dynamic client registration”, check that the config uses the native HTTP transport (type "http") with the Costory URL.
Allocation only works at project/key grain; one project per team is the practical split.
Set up in about 30 minutes
1. Connect OpenAI to Costory. Follow the OpenAI setup guide in the docs.
2. Map spend to owners. Use Virtual Dimensions to group accounts, projects, labels and tags into teams, products and environments, even where tagging is incomplete.
3. Add Costory to Claude Code.
Run in your terminal:
Then run /mcp inside Claude Code and select Authenticate to log in with your Costory account.
Full client instructions: MCP setup guide.
Good to know
Allocation only works at project/key grain; one project per team is the practical split.
FAQ
Does Claude Code get write access to my OpenAI account?
No. Costory reads billing data you export to it. MCP actions create Costory objects (alerts, reports, dashboards, events) and never change your cloud resources.
What OpenAI permissions does Costory need?
Read access to billing data only. The exact permissions are listed in the OpenAI setup guide.
Where is the data stored?
Costory stores and processes data in the EU.
How much does it cost?
Plans start at €250/month (up to €10M annual cloud spend), billed annually, with a 30-day free trial. See pricing.
Playbooks
Playbooks for OpenAI
What the Costory agent can watch, investigate and report on for you, with this in the loop.
All playbooks
Costory notices
A feature's model cost per request rises, or a new model ships.
Choose LLM models and routing based on cost per feature
The agent tracks per-model spend for each AI feature across OpenAI, Anthropic, Bedrock and Vertex AI, and recommends where a cheaper model or a routing change could fit.
See the playbook
Costory notices
A shared API or platform bill lands without an owner.
Split shared API costs across teams with usage data
Allocate untagged or shared API costs, like a single LLM key or a shared data platform account, to teams based on actual usage.
See the playbook
Costory notices
Every week, across every connected provider.
One showback across AWS, Google Cloud, Azure, AI and data platforms
Teams see all their costs in one place, whatever the provider: cloud, Kubernetes, data platforms and LLM APIs.
See the playbook
Costory notices
Daily, as LLM provider costs refresh.
LLM cost allocation by team, feature and customer
Allocate model spend from OpenAI, Anthropic, Azure OpenAI, Bedrock, Vertex AI and OpenRouter to teams, features and customers.
See the playbook
Connect in under 30 minutes. Get your first optimization recommendation the next day.
30-day free trial. Copilot, €250/month, or Copilot Pro, €1,000/month, billed annually.
First optimization recommendation
The next day
ROI
Pays for itself with the first recommendation you act on
Setup
About 30 minutes
Data residency
EU
Access
Read-only on your cloud