RELATED 01
Finance Toolkit MCP Server
Calculates financial ratios, valuation models, risk metrics and macro series from raw statements
5.5kGitHub stars
34k/moPyPI downloads
86popularity score
Finance Database MCP Server
Screen and look up 300,000+ equities, ETFs, indices, currency pairs and crypto by classification
Official · Jeroen Bouma Python Remote (hosted) · Streamable HTTPLocal · stdio Finance and payments
85 popularity score ★ 9.6k ↓ 20k/mo
Connect your client
Remote (hosted)
Add in one click
Opens the app, which asks before it adds the server.
Or add it by command or config
Claude CodeTerminal
claude mcp add --transport http finance-database https://financedatabase.jeroenbouma.com/mcpClaude (desktop and web)Settings → Connectors → Add custom connector
https://financedatabase.jeroenbouma.com/mcpChatGPTSettings → Apps → Advanced → Developer mode, then Create app
https://financedatabase.jeroenbouma.com/mcpCursor~/.cursor/mcp.json
{
"mcpServers": {
"finance-database": {
"url": "https://financedatabase.jeroenbouma.com/mcp"
}
}
}VS Code.vscode/mcp.json
{
"servers": {
"finance-database": {
"type": "http",
"url": "https://financedatabase.jeroenbouma.com/mcp"
}
}
}Codex~/.codex/config.toml
[mcp_servers.finance-database]
url = "https://financedatabase.jeroenbouma.com/mcp"Gemini CLITerminal
gemini mcp add --scope user --transport http finance-database https://financedatabase.jeroenbouma.com/mcpLocal (PyPI)
Add in one click
Opens the app, which asks before it adds the server.
Or add it by command or config
Claude CodeTerminal
claude mcp add finance-database -- uvx --from 'financedatabase[mcp]' financedatabase-mcpClaude Desktopclaude_desktop_config.json
{
"mcpServers": {
"finance-database": {
"command": "uvx",
"args": [
"--from",
"financedatabase[mcp]",
"financedatabase-mcp"
]
}
}
}Cursor~/.cursor/mcp.json
{
"mcpServers": {
"finance-database": {
"command": "uvx",
"args": [
"--from",
"financedatabase[mcp]",
"financedatabase-mcp"
]
}
}
}VS Code.vscode/mcp.json
{
"servers": {
"finance-database": {
"type": "stdio",
"command": "uvx",
"args": [
"--from",
"financedatabase[mcp]",
"financedatabase-mcp"
]
}
}
}Codex~/.codex/config.toml
[mcp_servers.finance-database]
command = "uvx"
args = ["--from", "financedatabase[mcp]", "financedatabase-mcp"]Gemini CLITerminal
gemini mcp add --scope user finance-database uvx -- --from 'financedatabase[mcp]' financedatabase-mcpSettings 1
FD_MCP_ANALYTICS |
Optional: Set to 1 on a self-hosted HTTP server to publish anonymous usage totals at /stats. |
|---|
What it does
This server puts the open Finance Database, a community-maintained catalogue of more than 300,000 listed instruments, behind MCP tools. It answers questions such as which large-cap chipmakers are headquartered in a given country, which bond ETFs an issuer lists on US exchanges, or which tickers share one ISIN across exchanges and currencies.
There is one tool for each asset class (equities, ETFs, indices, currencies, cryptocurrencies), each filtering on fields like country, sector, industry, exchange, market cap tier, ETF category or issuer, plus a free-text query on symbol and name. Three more tools search by ticker, name, ISIN, CUSIP or FIGI across every class, list the valid values of a filter, and describe the asset classes. Filters take several comma-separated values and accept common synonyms and unaccented spellings; a value that does not match comes back with a suggested correction.
It can be used two ways: a free hosted endpoint run by the author that needs no account, or a local stdio process started through uvx, which downloads the database, caches it and checks for updates at most once a day. The same package can also serve over streamable HTTP for a team, and the repository ships a Dockerfile and docker-compose file for that. The data holds classifications and identifiers only; the author pairs it with the separate Finance Toolkit MCP server for statements, ratios and prices.
What it needs
| To run it yourself | Python |
|---|
Hosted: nothing beyond an MCP client that can add a remote HTTP server. Local: uv (uvx), which fetches a suitable Python and the financedatabase package with its mcp extra itself, plus network access on first use to download the database. No API key or account in either case.
Limits
Holds classifications and identifiers only: no prices, quotes or financial statements. Mutual funds and money market funds were dropped from the server in October 2026 because no public source keeps them current; exchange-traded funds remain under ETFs. Results return 25 rows by default and at most 200 per call, with paging. The database is refreshed weekly from public sources, so counts change and classifications can be wrong until corrected on GitHub. Clients that cannot start local processes (claude.ai, ChatGPT) must use the hosted endpoint or a self-hosted URL.
Who it suits
Suits analysts, investors and developers who want an assistant to build a complete, repeatable list of instruments from filters, or to resolve ISINs and tickers to listings, without an API key. Since it is read-only and needs no sign-in, the hosted endpoint is the quickest start; pick the local uvx process if queries should not leave your machine. It does not price or analyse anything: for financial statements, ratios, valuation and risk on the tickers it returns, the author's Finance Toolkit server (finance-toolkit) is the companion. For live quotes or trading, choose a market-data or broker server instead.
Tools 8
equities- Filter stocks by country, sector, industry, exchange, currency, market or market cap tier
etfs- Filter ETFs by category group, category, issuer family, currency or exchange
indices- Filter market indices by category group, category, currency or exchange
currencies- List currency pairs by base and quote currency
cryptos- List crypto pairs by coin and quote currency
search_instruments- Find a symbol by ticker, name, ISIN, CUSIP or FIGI across all asset classes
show_options- List the valid values for a filter, optionally narrowed by other filters
search_categories- Describe each asset class with its size and available filters