Field-by-field comparison

Compare the evidence before choosing an MCP.

Each signal stays separate, missing facts remain visible, and the column order follows your selection.

MCPs in this comparison

Finance ToolkitPayPal MCP

Uncheck an MCP to remove it. Comparing two or three keeps each signal in its own row.

Comparable alternatives

Azure MCP ServerMCP100 SelectedChrome DevTools MCPMCP100 SelectedContext7 MCPMCP100 SelectedFirecrawl MCPMCP100 SelectedGrafana Cloud MCPMCP100 SelectedGoogle MCP Toolbox for DatabasesMCP100 Selected

Shown only when capability or industry context overlaps.

Side-by-side MCP evidence comparison
Evidence field	Finance ToolkitWatchlist candidate	PayPal MCPMCP100 Selected
Task fit
Best for	Comparing companies, financial statements, ratios, performance, risk, models, market data, and economic indicators through documented financial metrics Analysts who can independently validate upstream data and calculation output before using it in a financial decision	Querying invoices, orders, subscriptions, disputes, and shipment information Sandbox prototyping of merchant workflows Teams able to issue, scope, rotate, and monitor PayPal credentials
Not ideal for	Brokerage, trade execution, portfolio mutation, or other transaction workflows outside the documented read-and-calculate tool surface Decisions that require guaranteed upstream data accuracy, complete hosted-service retention documentation, or offline-only processing	Unattended production payment, refund, or invoice decisions Users without PayPal developer credential management Assistants that only need public PayPal documentation
Avoid when	You cannot provide a dedicated Financial Modeling Prep API key to either the hosted OAuth flow or a controlled local configuration An AI-generated financial interpretation could be acted on without independent review of the source data, plan limits, and calculation assumptions	Production financial actions would execute without explicit confirmation Access tokens must be embedded in committed configuration Merchant-account activity cannot be monitored and credentials cannot be revoked quickly
Provenance
Provenance details	First-party MCP Jeroen Bouma · Publisher source ↗	First-party MCP PayPal · Publisher source ↗
Maintenance
Maintenance details	Repo: Aug 13, 2026 Package: Jul 14, 2026	Repo: Not documented Package: Not documented
Popularity evidence
30-day package downloads	41,272 PyPI Stats package downloads · checked 2026-08-14T18:46:26.000Z	No reliable download data
Shared repository stars	5,222 Shared repository stars (not MCP-specific) · checked 2026-08-14T18:46:26.000Z	Not documented
External adoption evidence	github: repository stars: 5,222Non-comparable contextual signal.github: repository stars · Admission snapshot · checked 2026-08-14T18:46:26.000ZEvidence source ↗ PyPI Stats: package downloads: 41,272Non-comparable contextual signal.PyPI Stats: package downloads · Provider last month · checked 2026-08-14T18:46:26.000ZEvidence source ↗	Not documented
Client coverage
Client coverage details	Claude Code✓ Streamable HTTP Claude Desktop✓ Streamable HTTP Codex✓ Streamable HTTP VS Code✓ Streamable HTTP Cursor✓ Streamable HTTP OpenCode✓ Streamable HTTP	Claude Code✓ Local stdio Claude Desktop✓ Local stdio Codex✓ Local stdio VS Code✓ Local stdio Cursor✓ Local stdio OpenCode✓ Local stdio
Client coverage note	Local stdio runs on your computer. Streamable HTTP connects to a remotely hosted MCP server.
Access and data
Authentication	The hosted endpoint uses an OAuth consent flow that asks for a Financial Modeling Prep API key on first use. Local stdio uses FINANCIAL_MODELING_PREP_API_KEY or FINANCETOOLKIT_ENV_FILE; inline configuration takes precedence when both are present. Use a dedicated key and remove the MCP entry, local secret, and publisher authorization when access is no longer needed.	Local mode uses PAYPAL_ACCESS_TOKEN and PAYPAL_ENVIRONMENT. Remote mode supports PayPal login and consent or token authorization; PayPal publishes separate Sandbox and Production endpoints.
Cost	The MIT-licensed Finance Toolkit package and MCP documentation list no separate MCP usage fee. The documented Financial Modeling Prep free plan is limited to 250 requests per day, five years of data, and US-listed companies; premium data and some metrics require a paid FMP plan, and normal hosting, client, and model costs may also apply.	The MCP package is publicly available. Normal PayPal account, transaction, product, API, and merchant-service fees or limits remain applicable.
Permissions	The documented MCP tools retrieve company, market, risk, model, and economic data and calculate financial metrics with the operator's Financial Modeling Prep key. The documented surface does not include brokerage or trade execution. Local setup writes client configuration, and the local server writes a SQLite result cache in the configured Finance Toolkit data directory.	Capabilities depend on enabled tools and the PayPal credential and can include reads and mutations across invoices, orders, subscriptions, disputes, shipments, transaction details, and related merchant resources.
Data handling	The hosted service receives the Financial Modeling Prep API key through its OAuth consent page, remembers it, fetches live financial data, and returns results to the connected MCP client; hosted retention and logging are not documented. Local mode reads the key from inline environment configuration or a local environment file and caches result DataFrames in SQLite with a documented default 600-second TTL. Finance Toolkit may fall back from Financial Modeling Prep to Yahoo Finance unless the source is enforced.	The local MCP process sends merchant, customer, payment, and task data to PayPal APIs and returns results to the MCP client. The hosted option processes requests through PayPal's remote MCP service.
Limitations
Tradeoffs	The hosted endpoint is simple to connect through OAuth, but it receives and remembers the Financial Modeling Prep API key while its retention and logging controls are not documented on the MCP page. The local server keeps the MCP process under operator control, but stores configuration locally and caches returned DataFrames in SQLite while still retrieving data from external providers. More than 200 transparent calculations improve analytical coverage, but availability and accuracy still depend on Financial Modeling Prep plan limits and the documented Yahoo Finance fallback.	The local npm server is portable and publisher documented, but executes package code on the user's machine. The documented `--tools=all` setup is convenient but exposes a broad commerce surface; narrower tool selection is safer where supported. Sandbox reduces financial impact but does not replace production authorization and operational controls.
Risk context	Protect the Financial Modeling Prep API key, prefer a dedicated credential, and independently verify material financial output. The publisher disclaims responsibility for upstream data accuracy and documents an automatic Yahoo Finance fallback unless a source is enforced. Hosted key retention and logging are not documented, while local mode leaves configuration and cached financial results on disk.	Production access can create financial loss or customer-facing changes and can disclose transaction data. Start with Sandbox, keep credentials out of files and history, minimize tools, and require approval for every mutation.
Evidence date
Editorial review	2026-08-14	2026-07-24
Candidate evidence	2026-08-14T18:46:26.000Z	2026-07-29T20:58:26.000Z

Popularity, maintenance, fit, permissions, and client support are independent evidence fields.

Read the evidence method →