azure_price_searchLABSCO SUMMARY
It answers the region question, which is where Azure bills actually diverge.
Searching and comparing go straight at the Azure Retail Prices API, so the same SKU can be priced in two regions in one pass and in the currency you work in. The estimate call turns that into a monthly figure for the hours you will really run.
What it is
A pricing server over the Azure Retail Prices API, run locally from a checkout. Every figure comes from the live API rather than a stored table.
What you get
- azure_price_search — Azure retail prices with flexible filtering, in the currency you ask for
- azure_price_compare — the same service priced across regions and SKUs
- azure_cost_estimate — an estimated cost for a usage pattern you describe, rather than an always-on assumption
- azure_discover_skus — the SKUs available for a service
- azure_sku_discovery — SKU discovery with fuzzy matching, for when you can only describe roughly what you want
- Savings plan pricing wherever the API publishes it
Requirements
Python and a clone of the repository — the setup script builds the environment, and the client config points at the module. The config carries no credentials.
Setup effort
Build from source — git clone https://github.com/charris-msft/azure-pricing-mcp.git
USE CASES
01
Deciding which region to deploy in
Price the same service across the candidate regions in one call, instead of reading the pricing page twice.
02
Putting a number on a proposal
Describe the usage pattern and get a cost you can defend, with the underlying prices available behind it.
03
Finding the right VM size
Describe what you need in words, let fuzzy matching find the SKUs, then price the shortlist.
04
Checking whether a workload is worth turning off at night
Estimate the same service at part-time hours and compare it to always-on before building the automation.
5 tools
Listed in project docs. Your client may show a different list.
azure_price_compareazure_cost_estimateazure_discover_skusazure_sku_discoveryBEFORE YOU INSTALL
What to know first
Tools you get
5 tools named in the project's documentation.
Requires
Python with pip
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
Server setup
TERMINAL
git clone https://github.com/charris-msft/azure-pricing-mcp.git
cd azure-pricing-mcp
pip install -r requirements.txt
python -m azure_pricing_server --testMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"azure-pricing": {
"command": "python",
"args": [
"-m",
"azure_pricing_server"
],
"cwd": "<YOUR_PATH_TO_AZURE_PRICING>"
}
}
}<YOUR_PATH_TO_AZURE_PRICING> is your clone's folder. If you installed the requirements into a virtual environment instead (the README's setup script and manual setup both create .venv), set command to that environment's Python — .venv/bin/python, or .venv\Scripts\python.exe on Windows.
