MostExpensiveWatches — watch auction records API, dataset, MCP server & CLI

Documented ultra-luxury watch auction data ($100k–$31M realized prices, primary-sourced), free to use, built for AI agents, CLIs and apps. No API key required.

Live site: https://mostexpensivewatches.net

For AI agents — MCP server (one line of config)

{"mcpServers": {"mostexpensivewatches": {"url": "https://mostexpensivewatches.net/mcp"}}}

Streamable HTTP, JSON-RPC 2.0. Tools: search_inventory, lookup_reference, search_auction_comps, list_upcoming_auctions, get_market_estimate, get_brand_info, search_answers, get_site_stats. Install docs: https://mostexpensivewatches.net/mcp/install

Related MCP server: AI Makerspace MCP Server

Datasets (CC-BY 4.0, schema.org Dataset, refreshed automatically)

Snapshots in this repo are point-in-time copies; the live endpoints recompile daily via an autonomous verification pipeline. Cite as "MostExpensiveWatches Record Ledger".

API (43 endpoints, no key)

• OpenAPI 3.1: openapi.json · live: https://mostexpensivewatches.net/openapi.json

• Docs: https://mostexpensivewatches.net/api

• One-call snapshot: https://mostexpensivewatches.net/api/snapshot.json

• Examples: https://mostexpensivewatches.net/api/examples.json

• llms.txt: https://mostexpensivewatches.net/llms.txt

curl -s "https://mostexpensivewatches.net/api/listings/search?q=nautilus&limit=3"
curl -s "https://mostexpensivewatches.net/api/auctions?brand=Rolex&limit=5"
curl -s "https://mostexpensivewatches.net/api/answers/patek-5711-market-price.json"

CLI (dependency-free Python)

curl -sO https://mostexpensivewatches.net/cli/mew.py
python3 mew.py search "daytona" --limit 5

Also in this repo: mew.py

Human surfaces the data powers

• The Record Ledger — every auction record, documented

• How much is it worth? — 650 reference valuations

• HAMMER — daily guess-the-hammer-price game

• Grail Index · Price Index · Weekly brief

License

Data: CC-BY 4.0 (attribution: MostExpensiveWatches). Code (mew.py): MIT.