How to find company earnings with ChatGPT, Claude, or any AI assistant

Finding a company's earnings means tracking down when it reports, what it reported and what management said on the call. Companies publish all three across their investor relations site, SEC filings and a webcast. An AI assistant connected to Firecrawl Alexandria can pull them in one request from Fiscal.ai, a licensed financial data provider, with every number linked to its source filing.

Paste this into any assistant connected to Firecrawl and change the company:

Use Firecrawl Alexandria to find the latest earnings for [company or
ticker] from Fiscal.ai. Show me the price of each tool before you run it.
 
Give me:
- The next earnings date, the session (before or after market) and
  whether the date is confirmed
- Last quarter's revenue, net income and diluted EPS, with the same
  quarter a year earlier for comparison
- Three things management stressed on the earnings call
- A link to the filing behind each number
 
Quote figures exactly as reported. Don't give investment advice.

"Earnings" usually means one quarterly or annual report and everything that comes with it:

• The date. When the company releases results, and when it holds the call.
• The reported numbers. Revenue, operating profit, net income and earnings per share (EPS), as they appear in the income statement.
• Adjusted numbers. Management's non-GAAP figures, which leave out items such as stock compensation. Analyst estimates are usually quoted on this basis.
• The call. Management's prepared remarks and the analyst Q&A, where guidance and context come out.

1. Connect Firecrawl to your assistant. Claude, ChatGPT, Cursor, Claude Code and any other MCP client work; setup for each is below.
2. Name the company. A ticker with its exchange (NVDA on NASDAQ) is the most precise. A company name works too: the assistant resolves it to an identifier first.
3. Ask for the date, the numbers or the call. The assistant finds the matching Fiscal.ai tool in Firecrawl Alexandria, shows its price and runs it.
4. Check the sources. Each figure in Fiscal.ai's statements carries a link to the filing page it came from, so you can open the original before you use it.

Six Fiscal.ai tools in Firecrawl Alexandria cover the earnings cycle:

As an example, a test run for Nvidia on Nov 9, 2026 returned a confirmed date of Nov 17, 2026, after market close, for its fiscal Q3 2027 release and call. The income statement for the quarter ended Jul 26, 2026 showed revenue of $96.2 billion and diluted EPS of $2.46, each linked to the 10-Q page it came from, with quarterly history back to fiscal 2015.

1. Open the Firecrawl connector and select Connect.
2. Sign in to Firecrawl, pick a team and approve access. You can create a free account in the same step.
3. In a chat, open the + menu, go to Connectors and turn on Firecrawl.

Then ask in plain words:

Use Firecrawl Alexandria to check when Microsoft reports next and pull
its last four quarters of revenue and diluted EPS from Fiscal.ai.

1. Install the Firecrawl plugin from the ChatGPT plugin directory.
2. Sign in to Firecrawl when ChatGPT asks, or sign up first.
3. Start your message with @firecrawl so ChatGPT routes it through the plugin.

@firecrawl Get Apple's earnings call transcript for its latest quarter
from Fiscal.ai and list the questions analysts asked about margins.

Any agent that supports MCP can use the Firecrawl MCP server, Firecrawl Alexandria included. Add the server, then sign in to Firecrawl in the browser window that opens.

• Claude Code: run claude mcp add --transport http firecrawl https://mcp.firecrawl.dev/v2/mcp-oauth, then finish signing in from /mcp.
• Cursor or any client with a JSON config: add the server below to your MCP settings.

{
  "mcpServers": {
    "firecrawl": {
      "url": "https://mcp.firecrawl.dev/v2/mcp-oauth"
    }
  }
}

The agent gets three tools: firecrawl_search to find providers, firecrawl_find_tools to read a tool's inputs (and its price, when called with includeCosts: true), and firecrawl_scrape to run it. Ask it the same way you would ask Claude or ChatGPT.

To put earnings data into your own app or spreadsheet, call the Fiscal.ai tools directly with the Firecrawl SDK (pip install firecrawl-py):

from firecrawl import Firecrawl
 
app = Firecrawl(api_key="fc-YOUR_API_KEY")
company = {"ticker": "NVDA", "exchange": "NASDAQ"}
 
# When does it report next? 15 credits
calendar = app.scrape(
    alexandria={
        "provider": "fiscal-ai",
        "capability": "events/calendar",
        "options": {**company, "pageSize": 5},
    }
)
 
# What did it report? Standardized quarterly income statement, 15 credits
income = app.scrape(
    alexandria={
        "provider": "fiscal-ai",
        "capability": "financials/income-statement",
        "options": {**company, "periodType": "quarterly"},
    }
)
 
latest = income.alexandria[0].data["data"][0]
print(latest["fiscalYear"], latest["fiscalQuarter"])
print(latest["metricsValues"]["income_statement_diluted_eps"]["value"])

Firecrawl Alexandria needs an API key; see the Firecrawl docs for the full reference.

Finding a tool and checking its price is free in every setup above. Running a tool uses Firecrawl credits at the price shown in the table, and your assistant reports which tool it ran and what it spent. A full earnings check (date, income statement and transcript) comes to 60 credits. Plans and credit prices are on Firecrawl pricing.

• Estimated dates move. Read the status on each calendar entry. Only a confirmed date is fixed.
• Coverage varies by company. Segment and KPI data is narrower than the main financial statements and depends on what each company reports.
• No investment advice. The tools return reported data. What you do with it is your call.

• How to use Claude for finance: a broader research workflow across statements, transcripts and economic data.
• Fiscal.ai in Firecrawl Alexandria: every Fiscal.ai tool, from ownership data to fund letters.
• Finance data providers: other financial sources your assistant can call.