Model Context Protocol · official server

Steam MCP server — your AI assistant, trained on the entire Steam API.

Connect it to Claude Code, GitHub Copilot, Codex, Cursor or any MCP client. Your coding agent looks up endpoints, understands Steam data concepts and runs safe read-only test calls — no docs tab, no hallucinated parameters.

Set up in 2 minutes ↓ https://mcp.steamwebapi.com/mcp

// what this looks like in practice

You: "Build me a CS2 inventory value widget."

Agent → search_endpoints("inventory cs2")

Agent → get_endpoint("/steam/api/inventory")

Agent → get_concept("pricereal")

Agent: "Done — fetching server-side with your key, summing pricereal per item."

Connect your editor

Pick your client, copy the snippet. It's a hosted URL — nothing to install.

Include API key (enables live test calls — step 2)

Add your API key (optional)

Docs, schemas and concepts work without any key. With a key — tick the checkbox above, or send X-Api-Key or Authorization: Bearer — the agent can also run live read-only test calls. Your key is in the dashboard, top-right in the "Your API Key" card.

Ask in plain language

Copy any of these into your editor's chat — the trace shows which tools the agent uses.

"Build a Node.js service that fetches a CS2 inventory and sums its market value in USD."

→ search_endpoints → get_endpoint → get_concept

"Which endpoint gives me price history for one item, and what does the response look like?"

→ search_endpoints → call_readonly_endpoint

"What is the difference between pricelatestsell and pricereal?"

→ get_concept

"We expect ~50,000 requests per day for item prices — which plan do we need?"

→ get_pricing

Kickstart a project

Complete starter prompts for common builds — copy one into your connected editor and the agent scaffolds the whole thing using the MCP tools. Every endpoint referenced below is checked against the live API before this page is built.

Inventory value tracker

Web app that loads any Steam inventory and shows its total market value.

Using the steamwebapi MCP server, build a web app that takes a Steam ID or profile URL, fetches the user's CS2 inventory via the inventory endpoint, and shows every item with image, wear and current pricereal, plus the total inventory value in USD. Look up the right endpoints and price field semantics via the MCP tools. Fetch server-side with my API key from an env var — never expose it in the browser.

or open in Claude Lovable Bolt.new ChatGPT

Price alert Discord bot

Bot that watches items and pings a channel when the price crosses a threshold.

Using the steamwebapi MCP server, build a Discord bot (Node.js) where users register an item by market hash name plus a target price. Poll the item endpoint on an interval that respects my plan's rate limits (check get_pricing), compare pricereal against each target, and post an embed with item image, current price and best marketplace when a threshold is crossed. API key from env var.

or open in Claude Lovable Bolt.new ChatGPT

Marketplace price comparison page

Item page comparing all marketplaces, like our Explorer.

Using the steamwebapi MCP server, build an item detail page for any CS2 item: header with image and rarity/wear badges, a table comparing every marketplace from the prices[] array (logo, stock, price, best-price highlight), Doppler phase variants with images, and a chart from latest10steamsales. Check get_concept for the difference between price* and pricereal* fields. Server-side fetching with my API key.

or open in Claude Lovable Bolt.new ChatGPT

Portfolio dashboard

Track your own items over time with daily snapshots and P/L.

Using the steamwebapi MCP server, build a portfolio dashboard: I enter items I own with my buy price, the app fetches current pricereal for each daily and stores snapshots in SQLite. Show a table with buy price, current price, profit/loss in $ and %, and a line chart of total portfolio value over time. Verify the endpoint response shape with call_readonly_endpoint before writing the parser. API key from env var.

or open in Claude Lovable Bolt.new ChatGPT

What the Steam MCP server lets your agent do

The actual 10 tools the server exposes. Your agent picks them automatically — you never call them by hand.

Discover endpoints

search_endpoints

Free-text search across every endpoint and concept: "inventory cs2", "float screenshot".

list_endpoints

Browse the full API grouped by tag, or one tag at a time.

get_endpoint

The complete OpenAPI spec for one path: parameters, types, required flags, response fields.

Understand the data

get_concept / list_concepts

Explains market hash names, float values, price fields, trade locks, auth — with aliases.

list_games

Every supported game with its slug for the game parameter.

list_markets / list_products

Third-party marketplaces we track and the product packages that cover them.

Decide & verify

get_pricing

Public plans, per-endpoint rate limits and a plan recommendation for your expected daily volume.

call_readonly_endpoint

Runs a documented GET with your key so the agent sees a real response shape. Refuses all writes.

Safe by design

The Steam MCP server can read documentation and run GET requests — nothing else. It refuses every write method and blocks the Proxy, Steam Guard and Tradeoffer endpoint groups entirely, so an agent can never touch trades or account state.

Read the API documentation Ask in Discord

GET-only: every write method (POST, PUT, PATCH, DELETE) is refused.

Proxy, Steam Guard and Tradeoffer endpoints are blocked entirely.

Your API key is read from the request header per call — never stored on the server.

Test calls count against your normal plan limits, nothing extra.

Prefer a different workflow?

Not every question needs an editor. These two work without any setup.

ChatGPT Custom GPT

Our custom GPT trained on the steamwebapi API. Quick answers to questions while you build.

Open custom GPT →

Documentation

Full API reference with a "Copy for Agent" button — every endpoint in agent-ready format.

Open documentation →

Resources the MCP server exposes

docs://swagger Live OpenAPI 3.0 specification (5-minute cache).
docs://llms-full Curated long-form overview with all games, markets, products, endpoint guidance and competitor context.
/docs/raw/endpoints.md LLM-ready endpoint knowledge layer for every public OpenAPI operation.
/docs/raw/market-price-api.md Market price, item history and CS2 market index endpoint guide.
/docs/raw/inventory-api.md Inventory, profile and account-data endpoint guide.
/docs/raw/trade-api.md Trade offer, Steam login and Steam Guard endpoint guide.
/docs/raw/float-api.md CS2 float, inspect-link and screenshot endpoint guide.
docs://markdown/<file> Static documentation files (price-doc, response-format, market-index history, …).
docs://markdown/phases.json CS2 Doppler phase image-hash table.

How it works

Your AI agent talks to the MCP server over Streamable HTTP. The server reads the live OpenAPI specification of steamwebapi.com so every endpoint, parameter and response stays in sync with the production API. Domain concepts (Doppler phases, wear ranges, pricing fields, assetid vs classid) are exposed as a curated glossary.

Discovery tools work without an API key. The optional read-only executor accepts your existing steamwebapi API key via the MCP Authorization or X-Api-Key header and only calls documented GET endpoints. Mutating Tradeoffer, Steam Guard and Proxy actions are blocked by design. Use get_pricing when the agent should pick a package, check endpoint quotas or explain when to upgrade.