mcpField
C
corpusAI Cloud Pricing

corpusAI Cloud Pricing

Remote
cloud.trycorpus.aiCloud & Infra0Updated 4w ago

About corpusAI Cloud Pricing

corpusAI Cloud Pricing is a stdio MCP server for cloud, GPU, and LLM pricing data, including price histories and index fixings. It offers free tools and per-call paid data access using USDC, with hosted-server and prepaid-key options also described.

Key features

  1. 01AWS, GCP, and Azure spot prices
  2. 02GPU rental quotes and index fixings
  3. 03LLM token prices and histories
  4. 04Free catalog and model tools
  5. 05Paid calls settled in USDC

Use cases

  1. 01Reviewing cloud spot-price history
  2. 02Comparing GPU rental quotes and index data
  3. 03Checking LLM token prices
  4. 04Retrieving pricing data with per-call USDC payment

MCP support

Exposes cloud, GPU, and LLM pricing tools to MCP clients over stdio. Paid tools use USDC per call; the documentation also describes a hosted MCP endpoint for prepaid-key access.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://cloud.trycorpus.ai/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

All endpoints
Streamable HTTPhttps://cloud.trycorpus.ai/mcpAuth required
Packages
npmcorpus-cloud-pricing-mcp0.1.6

Tools

35

This tool takes no parameters.

Details

From the project README

corpus-cloud-pricing-mcp

MCP server (stdio) for corpusAI Cloud Pricing: AWS, GCP and Azure spot price history at event resolution, on-demand and reserved list prices, GPU rental quotes from neo-clouds and marketplaces, LLM token prices, and daily index fixings per GPU and segment (CX.H100SXM.NEO, CX.H100SXM.SPOT, ...). Free tools need nothing. Paid tools pay per call in USDC over x402 through Circle Gateway, from a wallet you configure.

The API also takes MPP (USDC on Tempo) and prepaid keys. To use a prepaid key from an agent, connect to the hosted MCP server at https://cloud.trycorpus.ai/mcp instead of this package (see the docs).

Install

Claude Code:

claude mcp add corpus-cloud-pricing -e CORPUS_PAYER_KEY=0x... -- npx -y corpus-cloud-pricing-mcp

Claude Desktop, Cursor or any MCP client (mcp.json):

{
  "mcpServers": {
    "corpus-cloud-pricing": {
      "command": "npx",
      "args": ["-y", "corpus-cloud-pricing-mcp"],
      "env": { "CORPUS_PAYER_KEY": "0x...", "CORPUS_CHAIN": "base", "CORPUS_MAX_USDC": "0.10" }
    }
  }
}

Without CORPUS_PAYER_KEY the free tools work and the paid tools return the price and accepted networks instead of data.

Paying

Paid calls are settled gaslessly from a Circle Gateway USDC balance held by the payer wallet. Fund it once:

CORPUS_PAYER_KEY=0x... npx corpus-cloud-pricing-mcp balance
CORPUS_PAYER_KEY=0x... CORPUS_CHAIN=base npx corpus-cloud-pricing-mcp deposit 2.00   # needs USDC and a little gas on that chain

CORPUS_CHAIN takes a Circle Gateway chain name as the Gateway SDK spells it (base, ethereum, arbitrum, worldChain, hyperEvm, ...). CORPUS_MAX_USDC (default 0.10) refuses any single call priced above it. Prices: 0.01 USDC for single-type or single-fixing calls, 0.05 for full histories, 0.10 for region or market snapshots.

Tools

Free: catalog, instances, gpu_models, index_tickers, status, tokens_models. Paid: spot_daily, spot_hourly, spot_events, spot_latest, spot_snapshot, spot_monthly, ondemand, reserved, gpu_quotes, gpu_index, index_fixing, index_history, index_board, term_instance, term_curve, tokens_quotes, tokens_fixing, tokens_history, tokens_board, capacity_stress, capacity_history, capacity_board.

The hosted server generates its tools from the OpenAPI document, so its set is larger and a few names differ (its GPU term structure tool is term, where this package says term_curve).

Methodology for the index fixings: https://cloud.trycorpus.ai/methodology. Data license: https://cloud.trycorpus.ai/license. Contact: [email protected].

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.