Wisesheets API

What it is

Wisesheets sells two products. The spreadsheet add-in puts financial data in cells; this is the REST API it launched on 24 July 2026, with its own plans, its own key and its own MCP server. The same SEC-sourced fundamentals underneath, a different thing to buy.

Base URL https://api.wisesheets.io/v1/, Bearer token, twelve documented data endpoints plus /me and /health. Company search and profiles, end-of-day and live prices, dividends, a /financials query taking up to 100 tickers on GET and 1,000 on POST, full statements, and a catalogue of the 231 metrics. Every response uses one envelope — a data array and a meta object — and numeric values come back as strings, so a client parses them without float damage.

The pitch is traceability. Each value carries a source object naming the XBRL tag, the accession number and the filing date, plus a confidence flag that separates verified facts from computed ones. asReported=true returns the raw XBRL facts instead of the standardised metric, which is the parameter that matters if you are reconciling against a filing rather than screening across thousands of them.

Pricing

Free is 5,000 requests a month, 200 calls a minute and five years of history. Pro is $19 a month for 80,000 requests, 600 a minute, fifteen years and the bulk POST endpoints; Elite $39 for 200,000 requests, 1,500 a minute and full history. Both paid cards are marked Billed Annually and their buttons request plan=pro_annual and plan=elite_annual — no month-to-month rate appears anywhere on the page.

The tier that decides the most is the one a reader skims past. Free, Pro and Elite all carry an individual-use licence, and the vendor defines it by exclusion — not permitted if you are registered or qualified with a regulator, if you use it on behalf of a company or fund, if you use it in a profession, trade or business, or if you manage or advise on money for other people. That removes most of the audience an API is normally sold to, and the remedy is Enterprise, which is quoted.

Data & coverage

10,412 US stocks, which is every SEC filer and nothing else, with 230+ standardised metrics and 20+ years of history taken from EDGAR XBRL rather than bought from an aggregator. Options and forex are labelled "Soon" on the product page; Canada, Europe and Asia Pacific sit on a roll-out plan with no date. Quotes are 15-minute delayed at every tier.

Financials return JSON by default or pipe-delimited CSV, in three layouts — long for time series, wide for a company per row, grouped by period.

Integrations

The MCP server is first-party and hosted at mcp.wisesheets.io/v1/mcp over streamable HTTP, with a browser sign-in instead of a pasted key; an unauthenticated POST answered 401 AUTH_MISSING on 14 September 2026. There are official connectors for Claude, ChatGPT and Perplexity, and documented setup for Gemini CLI, Cursor and VS Code.

There is no SDK. The docs give examples in Python, JavaScript, cURL and Google Sheets Apps Script, all against the plain HTTP interface, and the obvious package names are unclaimed — wisesheets returns 404 on both PyPI and npm.

Limitations

• US SEC filers only. No non-US listings, no options, no futures, no forex, no crypto; the two marked "Soon" carry no date.
• The individual-use licence disqualifies most professional readers before the price matters.
• "Webhook on new filings" is a bullet on the product page and appears nowhere in the API reference, which documents no webhook endpoint at all.
• The freshness claim does not hold together. The same FAQ says median EDGAR-to-API latency is around 15 minutes and that P95 is under 15 minutes, which cannot both be true.
• Every error response points at docs.wisesheets.io for its explanation, and that host does not resolve.
• No real-time quotes at any price, and no month-to-month billing on the published plans.

Alternatives

Financial Datasets is the closest thing in shape — SEC fundamentals over REST with a first-party MCP server — but has no free tier, starting at a $20 credit pack. Financial Modeling Prep is cheaper per request and covers non-US markets, without the citation trail. Daloopa and Fiscal.ai go deeper into as-filed detail for readers who will pay institutional prices, and neither is constrained to an individual-use licence.