Earnings Calendar API
We built this because predicted earnings dates are unreliable. A company says "late October" on its IR page, an aggregator picks that up, and six weeks later the date quietly shifts to early November. Nobody updates the feed. If your model traded on the original date, you missed the move entirely.
viaNexus tracks earnings dates for thousands of US-listed companies and confirms them from the primary source: the SEC 8-K Item 2.02 filing. When a company actually reports earnings, the 8-K is the legal record. Our classifier reads the filing within minutes of the SEC acceptance timestamp, extracts the date, matches it to the company, and locks the record. No IR scraping, no third-party relay.
What the data looks like
Every record has seven fields:
{
"symbol": "AAPL-US",
"earningsDate": "2027-01-30",
"status": "confirmed",
"confidence": 0.99,
"marketSession": "aftermarket",
"filingDate": "2027-01-30",
"predictionSources": ["historical_pattern", "sec_8k"]
}
status tells you whether you are looking at a prediction or a confirmed fact. confidence is a numeric score (0 to 1) that reflects how much evidence supports the date. A prediction based solely on historical quarterly cadence sits around 0.6. A date confirmed by an actual 8-K filing jumps to 0.95 or higher. marketSession tells you whether the company reported before the open or after the close, so you know which session to watch.
The API call
GET https://api.blueskyapi.com/v1/data/core/earnings_calendar/AAPL-US?token=YOUR_TOKEN
Returns the most recent record. Add &last=8 for the last 8 quarters. Drop the symbol entirely and use &from=2027-01-27&to=2027-01-31 to pull every company reporting that week.
How the confirmation pipeline works
The pipeline runs in two stages. First, we generate a predicted date from historical filing patterns: if Apple has reported on the last Thursday of January for the past six years, we predict the same pattern forward. The predicted date goes live with a moderate confidence score.
Second, we watch the SEC EDGAR feed continuously. When Apple's 8-K Item 2.02 lands, our classifier parses it, confirms the earnings date, updates the record, and bumps the confidence. The gap between SEC acceptance and our record update is typically minutes.
This two-stage approach means you always have a date, even months out, but you also always know how trustworthy that date is. The confidence field is the signal. If you are building a trading calendar that only acts on confirmed dates, filter on status=confirmed. If you want the full forward calendar including predictions, take everything and weight by confidence.
Coverage
US-listed equities that file with the SEC: NYSE, NASDAQ, NYSE American. Thousands of symbols from mega-caps to small-caps. Symbol format is ticker plus exchange suffix (AAPL-US, MSFT-US, TSLA-US). If you work with ISINs or CUSIPs, our reference data endpoints handle the mapping.
Using it with an AI agent
Connect viaNexus to Claude through the MCP server and ask "when does Tesla report next?" The agent calls the earnings calendar, gets the live record, and answers from the API instead of from stale training data. It can also pull the consensus estimate, recent news, and historical price reaction in the same conversation, since those are all separate datasets on the same connection.
Setup: search for viaNexus in the Claude integrations directory and click Connect.
See it live
Try it now in Claude: Search for viaNexus in the Claude integrations directory, connect your account, and ask "when does Apple report earnings?" The agent pulls the live record and answers from the API. The MCP server guide walks through each tool.
Browse the data catalog: The viaNexus console shows every dataset, its schema, and field descriptions. The earnings calendar is under the CORE workspace. You can see field names, types, and sample values before writing any code.
Pricing
Included in any viaNexus plan that covers the CORE product. The 14-day free trial unlocks everything, no credit card. Full pricing in the catalog.
Related
- Stock Fundamentals API — pull the actual reported numbers once earnings land.
- Market Data API — historical prices to backtest earnings reactions.
- Financial Data for AI Agents — the full MCP toolset and how agents use it.
- MCP Server Guide — connection setup for Claude, Cursor, and other clients.