如何透過 Claude 查看公司收入來源：以 Apple 的地區、產品及利潤率為例

All outputs are real tool calls to the Equibles MCP server from 21 October 2026, trimmed.

"Apple sells iPhones" is true and not very useful. The questions that move the story are which parts are growing, which are shrinking, and which carry the margin.

Companies answer those in their own filings, in the segment and disaggregated revenue tables. Here's how I get an AI agent to pull them and read them without the usual mistakes.

Connect it first

In Claude, open Customize → Connectors → Discover, search for Equibles, click Connect to Claude and Authorize, then switch it on in a chat from + → Connectors. In ChatGPT: Plugins, search Equibles, Install plugin. In Claude Code, claude mcp add --transport http equibles https://mcp.equibles.com/mcp and then /mcp to sign in. The free plan covers 100 requests a day; other clients are in the setup docs.

1. Pull every breakdown the company reports

The prompt: "Using Equibles, show Apple's revenue breakdown for the last four fiscal years." The agent calls GetRevenueBreakdown, which returns one table per way Apple cuts its revenue: by reporting segment, by country, by product, plus operating income per segment. The values come from the dimensional XBRL Apple tags in its 10-K, latest restated figures.

By product, in billions:

2. Turn amounts into shares and changes

Raw amounts hide the mix shift. "For each line, give the change from FY2022 to FY2025 and its share of total revenue in both years."

                 FY22→FY25   Share FY22   Share FY25
Services           +39.7%      19.81%       26.23%
iPhone              +2.0%      52.11%       50.36%
Mac                -16.1%      10.19%        8.10%
Wearables          -13.5%      10.46%        8.58%
iPad                -4.3%       7.43%        6.73%

One line stands out. Total revenue grew $21.8 billion over the three years, and Services alone grew $31.0 billion. Everything outside Services and iPhone shrank.

3. Do the same by region

The segment table answers a different question. Apple's segments are regions:

                 FY22→FY25   Share FY22   Share FY25
Europe             +16.7%      24.12%       26.68%
Rest of Asia Pac   +14.7%       7.45%        8.10%
Japan              +10.5%       6.59%        6.90%
Americas            +5.1%      43.02%       42.86%
Greater China      -13.2%      18.82%       15.47%

Greater China went from $74.2 billion to $64.4 billion while every other region grew. Europe is now the second-largest region by a wide margin.

4. Check which segments carry the margin

GetRevenueBreakdown also returns operating income by segment where the filer tags it, and derives the margin. FY2025:

Greater China is shrinking, but not because it got less profitable: its margin held near 42% the whole time. Japan has been the most profitable region in all four years.

One trap here. The five segments add up to $175.7 billion of operating income, against $133.1 billion for Apple as a whole. The $42.6 billion gap is R&D and corporate costs that Apple does not allocate to regions. Segment margins are before those costs, so don't compare them with the company's overall operating margin.

5. Don't add rows together

The tool warns about this, and it's worth putting in the prompt anyway: rows within one table can overlap when a company tags a parent and its parts on the same axis. Apple's country table is a good example. Its "China" row is exactly the Greater China segment, $64,377 million, so mixing the two tables double-counts it. Use the consolidated total that each table carries.

The prompt I reuse

Using Equibles, break down <TICKER>'s revenue for the last four fiscal years:
1. Every table the company reports: segment, geography, product, segment income.
2. For each line: change over the period and share of total, first and last year.
3. Segment operating margins, and the gap between the segment sum and the
   consolidated total.
4. Never add rows; use each table's consolidated total.

What a revenue breakdown can't tell you

• Product profitability. Apple reports operating income by region, not by product, so the data can't say how much more profitable Services is than iPhone.
• Anything inside the year. These tables are annual only; quarterly segment numbers aren't in this dataset.
• Renamed lines. A segment the company renames shows up as a new row, so a gap can be a relabel rather than a drop.

The same tables are on each company's financials page, like Apple's.

Nothing here is investment advice.