# Veritas Data

> Veritas Data is a freemium MCP server for financial data. US earnings dates marked confirmed or estimated, each linked to its SEC or company source. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Starting price:** $19 a month
- **Platforms:** API, Web
- **Website:** https://veritasdata.dev
- **Repository:** https://github.com/finn858/veritas
- **MCP Registry name:** dev.veritasdata/earnings
- **Categories:** [financial data](https://ai-powered.com/categories/financial-data)
- **Listing updated:** 2026-11-19

## About

The README and registry entry describe a service that tracks report dates for US-listed companies worth $300M or more, roughly 3,200 names at the time of writing. Each date is labelled confirmed when the company has announced it, or estimated when it is projected from the company's filing history, and the record links to the press release or EDGAR filing behind it.

Beyond the calendar, the same data source covers results, XBRL fundamentals, insider trades and filing indexes, drawn only from SEC EDGAR and company announcements. Access is through a REST API or an MCP server that exposes the data to AI clients such as Claude, Cursor and ChatGPT, with tool descriptions that tell the model to treat estimates as estimates and to cite the source.

The product suits traders watching earnings around a watchlist, analysts who need dated sources for their notes, and developers building apps that display report calendars. The site says that the Builder plan allows redistribution of the data, and it states that it does not provide investment advice.

## Features

- **Confirmed or estimated status:** Every report date is marked as confirmed by the company or estimated from its filing cadence.
- **Source link on each record:** Dates and values link back to the press release, 8-K or EDGAR filing they came from.
- **Call time and webcast:** Confirmed events include the Eastern conference call time and webcast link from the announcement.
- **Change feed:** Lists dates that moved, were confirmed, conflicted or were reported, with timestamps.
- **MCP server access:** Remote streamable-HTTP MCP endpoint, authenticated with an API key or OAuth.
- **Watchlists and alerts:** Free tier includes one watchlist of 25 tickers and email alerts on up to 10 tickers.

## Good for

- Check when a list of US small and mid caps are due to report earnings
- Cite the source of an earnings date in a research note or newsletter
- Give an AI assistant a reliable earnings calendar it can reference with links
- Track dates that move or get confirmed during a reporting season
- Build an app that shows earnings dates to its own users under a Builder plan

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http earnings https://api.veritasdata.dev/mcp
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://api.veritasdata.dev/mcp
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "earnings": {
      "url": "https://api.veritasdata.dev/mcp"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "earnings": {
      "type": "http",
      "url": "https://api.veritasdata.dev/mcp"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "earnings": {
      "serverUrl": "https://api.veritasdata.dev/mcp"
    }
  }
}
```

### Grok Bot (Grok Bot desktop app → Connect apps → Add a custom connector)

```sh
https://api.veritasdata.dev/mcp
```

Enter a Name (for example "Veritas Data"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin.

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "earnings": {
      "type": "http",
      "url": "https://api.veritasdata.dev/mcp"
    }
  }
}
```

## Tools

17 of 18 tools say they're read-only. 1 of 18 tools says it can delete or change data. From the server's own tool annotations; not verified by us.

- `get_next_earnings` (says read-only): Next earnings report date for one US ticker: report_date, BMO/AMC timing, status, confidence, source_url. Always state the status: 'confirmed' is company-announced (cite source_url); 'estimated' is a projection, never present it as confirmed. Paid plans add `behaviour`: past announcement habits, not a forecast. as_of (ISO timestamp) replays the answer from the change history: date and status only.
- `list_earnings` (says read-only): Earnings events in a date window (max 120 days), filtered by status, cap_bucket (small, mid, large, mega), timing (BMO, AMC) or one of the caller's watchlists (name or id; see list_watchlists). Relay each event's status, cite source_url for confirmed dates, label estimates as projections. as_of replays the window from change history: date and status only, no other filters.
- `get_date_changes` (says read-only): Earnings dates that moved, were confirmed, went into conflict or were reported since an ISO timestamp, with old and new date and status and the source_url behind each; cite it. Filters: ticker, cap_bucket, watchlist. Oldest first; when has_more is true, call again with since = next_since.
- `get_earnings_history` (says read-only): Past earnings release dates for one ticker: weekday, BMO/AMC timing, days after quarter end, and the EDGAR filing behind each (source_url). Use it to judge how regular the company's cadence is.
- `get_data_coverage` (says read-only): How far back data goes for one ticker: release count, first and last release, a count per year, whether releases rest on 8-K Item 2.02 or 10-Q/10-K filings, and when tracking began. Use it before asking for long histories.
- `get_earnings_results` (says read-only): Earnings releases for one ticker: revenue and EPS (GAAP and non-GAAP) read by software from the press release, each with its sentence; guidance sentences; call time; exhibit URL. Figures can be wrong: quote the sentence, cite source_url, relay disagrees_with_xbrl and the accuracy block. A null carries a reason (not_stated, split_by_business, no_scale, not_read): report it, never fill the gap.
- `get_guidance` (says read-only): Guidance statements for one ticker by release, newest first (beta): the verbatim sentence and the exhibit URL. metric, low, high and period_guided are filled only when the sentence states metric, scale and period, otherwise null with a reason. Quote the sentence, cite the exhibit, verify any number against the sentence. Trader plans and up.
- `get_fundamentals` (says read-only): Financial statements for one ticker from its XBRL filings, per quarter or year: revenue, gross profit, operating and net income, diluted EPS and shares, cash flow, capex, free cash flow, cash, debt, assets, liabilities, equity. Each number names its us-gaap tag and accession; cite them. Nulls say why. period is quarterly (default) or annual.
- `get_fact` (says read-only): Raw time series of one us-gaap concept for a ticker (e.g. Revenues, NetIncomeLoss, EarningsPerShareDiluted, Assets), each value with period, form, filing date and accession. Use for exact filed figures; cite the accession. Unknown concepts return the list of stored ones.
- `get_insider_trades` (says read-only): Insider transactions for one ticker from Form 4 filings: owner, role, transaction code, shares, price, holdings after, footnotes, filing URL. Optional since date. around_earnings=true: trades within window_days (default 30) of each past report date. One row per reporting owner; group by accession and line before summing shares. Cite source_url.
- `get_filings` (says read-only): Filing index for one ticker from EDGAR: accession, form, filing and period dates, 8-K items, document URL. Filter by form (10-K, 10-Q, 8-K, 4, ...) and a since date. Use it to find the exact filing to cite.
- `search_filings` (says read-only): Ranked full-text search over sections extracted from 10-K and 10-Q filings (mda, risk_factors, results_of_operations, liquidity). Each hit carries accession, section, a snippet and source_url; cite them. Filters: ticker, form (10-K or 10-Q), since date. Coverage starts 2025-01-01 and grows nightly.
- `get_filing_sections` (says read-only): Sections extracted from one 10-K or 10-Q, by accession (0000320193-26-000008): names, headings, sizes and extraction status. Pass section (mda, risk_factors, results_of_operations, liquidity) for that section's clean text. Cite source_url, the filing's primary document on EDGAR.
- `get_section_diff` (says read-only): What changed in one section between a company's two most recent filings of the same form: character counts, a similarity ratio and the changed passages. Both filings are cited (accession, source_url). section: mda, risk_factors, results_of_operations, liquidity. form: 10-K or 10-Q; default, the newest extracted filing's form.
- `get_corporate_events` (says read-only): Corporate events read from stored filings, newest first: 8-K items with one unambiguous meaning (material_agreement, acquisition_completed, auditor_change, restatement, officer_change) and late_filing from NT 10-K/10-Q. Every row cites its filing; relay source_url. Filters: types (comma-separated), ticker, since date.
- `get_congress_trades` (says read-only): US House members' reported stock trades: STOCK Act periodic transaction reports from the Clerk's public files. Amounts are the bands the filings state, never exact figures, and members report up to 45 days late. Filters: ticker, member last name, since date. Cite source_url on every row. Disclosure, not advice.
- `list_watchlists` (says read-only): The caller's watchlists with their tickers. Watchlists belong to the account behind the API key; pass a list's name (or id) as the `watchlist` argument of list_earnings or get_date_changes to answer questions like 'what on my list reports this week'.
- `set_watchlist` (says it can change data): Create the named watchlist, or replace its tickers if it exists. Tickers outside the covered universe are kept and reported in `unknown`. Free accounts: 1 list of 25 tickers; paid plans: 5 lists of 500.

## FAQ

### Where does the data come from?

The site says it uses SEC EDGAR filings such as 8-Ks, 10-Qs, 10-Ks, Form 4 and XBRL, plus announcements companies publish themselves. It states that no data is bought from a vendor.

### What does an estimated date mean?

It is a projection from the company's past reporting pattern, with a confidence score from 0 to 1. The site says it is not a company statement.

### Is there a free tier?

Yes. Signing in with GitHub or Google gives a key with 100 calls in total across every endpoint, with no card required.

### Which companies are covered?

Operating companies listed on NYSE, Nasdaq or NYSE American with a market cap of $300M or more. Preferreds, warrants, ETFs, SPACs and similar securities are not yet covered.

### Is Veritas Data free?

Veritas Data has a free plan, with paid plans for more usage or features. Paid plans start at $19 a month.

### How do I add Veritas Data to Claude Code?

Run this command in your terminal: claude mcp add --transport http earnings https://api.veritasdata.dev/mcp

### What are the best alternatives to Veritas Data?

The closest alternatives to Veritas Data are Companies Daily – SEC filings of US companies, US Company Financials (SEC XBRL) MCP, SEC Filing Watch, FinanceQuery and Market Pi-Rates.

## Where this comes from

- Description: written from the project's website and its MCP Registry entry.
- Status and tools: checked automatically by AIPoweredBot on 20 November 2026.
- Install steps: from the official MCP Registry entry.

## Alternatives

- [Companies Daily – SEC filings of US companies](https://ai-powered.com/mcp/companies-daily-sec-filings-of-us-companies): Companies Daily – SEC filings of US companies is a free MCP server for legal and government.
- [US Company Financials (SEC XBRL) MCP](https://ai-powered.com/mcp/us-company-financials-sec-xbrl-mcp): US Company Financials (SEC XBRL) MCP is a freemium MCP server for finance and payments.
- [SEC Filing Watch](https://ai-powered.com/mcp/sec-filing-watch): SEC Filing Watch is a freemium MCP server for legal and government.
- [FinanceQuery](https://ai-powered.com/mcp/financequery): FinanceQuery is an open-source MCP server for finance and payments.
- [Market Pi-Rates](https://ai-powered.com/mcp/market-pi-rates): Market Pi-Rates is a freemium MCP server for finance and payments.
- [Vigil by Reno Labs](https://ai-powered.com/mcp/vigil-by-reno-labs): Vigil by Reno Labs is an MCP server for financial data.
- [Finance: Stock, ETF & Crypto Data](https://ai-powered.com/mcp/finance-stock-etf-and-crypto-data): Finance: Stock, ETF & Crypto Data is a paid MCP server for finance and payments.
- [US Insider Trades (SEC Form 4) MCP](https://ai-powered.com/mcp/us-insider-trades-sec-form-4-mcp): US Insider Trades (SEC Form 4) MCP is a freemium MCP server for financial data.

---
Source: https://ai-powered.com/mcp/veritas-data
