REST + MCP for financial agents

The financial data layer for AI agents

Ground agents in structured US, HK, A-share, filings, financials, ownership, and news data through predictable APIs and machine-readable docs, with crypto coverage on the roadmap.

agent.run.market_context

Agent request

Analyze margin risk for NVDA, Tencent, and Kweichow Moutai with sourced market context.

US
1GET /filings/sec?market=us&ticker=NVDA&form_type=10-K

Extract risk factors and management discussion.

HK
2GET /financials/income-statements?market=hk&ticker=00700

Normalize reported revenue and operating margins.

A-share
3GET /company/announcements?market=CN&ticker=600519

Retrieve recent exchange announcements.

Grounded context

Returned structured fields, ranked source excerpts, and market-specific notes without stuffing whole filings into the prompt.

OpenAPI schemaMarket filterSource linksMCP-ready

First workflows

Financials / filings / prices / news

Start with the tasks investors already ask agents to run.

Agent contract

REST + OpenAPI + llms.txt

Give app backends and agent runtimes the same discovery surface.

Source context

Source-aware responses

Return structured fields with links, report dates, and market notes.

Next coverage

Ownership / transcripts / estimates / KPIs

Show roadmap data separately from what can be tested today.

Agent workflows

Designed for repeatable financial research jobs.

Package market-aware data retrieval into workflows your app can schedule, audit, and hand to an agent runtime.

Ready

Earnings monitor

Track filings, income statements, price reaction, and news around reporting windows.

financials + filings + prices + news

Ready

Risk extractor

Pull risk-factor changes and source excerpts before a model drafts the summary.

SEC filings + company announcements

Ready

Multi-market watchlist

Normalize coverage across US, HK, and A-share symbols without custom glue, with crypto coverage tracked as roadmap.

market filters + endpoint catalog

Preview

Credit-aware automation

Keep recurring agent runs bounded by API keys, credits, and usage visibility.

dashboard + billing state

Key benefits

Grounding, discovery, and compact context.

01

Grounding

Stop financial hallucinations.

Agents get typed fields, source links, and market-specific context instead of guessing from stale text.

02

Discovery

Find data across markets.

Filter docs and endpoints by US, HK, A-shares, or roadmap crypto coverage so builders only see what applies.

03

Lower context

Cut prompt stuffing.

Retrieve the fields, tables, and excerpts the task needs before sending context to a model.

How it works

From question to grounded answer in three moves.

01

Connect

Use REST, MCP, OpenAPI, or SDK examples from the same contract-first API surface.

02

Retrieve

Select a market, dataset, and symbol. The API returns normalized JSON with source metadata.

03

Decide

Feed compact, cited financial context into research copilots, monitoring agents, or apps.

Use cases

Copy a prompt. Inspect the calls. Build the workflow.

Every scenario maps to a concrete endpoint sequence so developers can move from idea to API call quickly.

A-share

A-share research agent

Analyze financials, prices, news, and announcement roadmap coverage for one listed company.

Analyze 600519 recent financials, price trend, news, and disclosure roadmap coverage

HK

HK announcement monitor

Track company facts, financials, news, and HKEX announcement roadmap coverage for Hong Kong equities.

Summarize 00700's latest financials, news, and disclosure roadmap coverage

US

SEC filing analyst

Extract risk factors, income statement trends, insider activity, and source-backed context.

Extract margin risk from NVDA's latest 10-K and income statements

Multi-market

Portfolio monitoring agent

Watch holdings across US, HK, and A-shares with market-aware data availability; keep crypto in roadmap plans.

Monitor my portfolio for filings, disclosure roadmap coverage, price moves, and major news

Market data store

One catalog. Market-aware coverage.

Docs and request previews should adapt to the market a user cares about. A-share builders should see A-share endpoints first.

US

US equities, SEC filings, fundamentals, ownership, prices, and market news.

AAPLMSFTNVDA

HK

Hong Kong equities, financials, prices, news, and HKEX announcement roadmap coverage.

007000998803690

A-shares

Mainland China A-shares, prices, financials, news, and exchange announcement roadmap coverage.

600519000001300750

Crypto

Crypto coverage is on the roadmap; current public catalog focuses on equities.

BTC-USDETH-USDSOL-USD

Prices

GET /market/bars

US

Full

HK

Full

A-shares

Full

Crypto

Planned

Company facts

GET /company/facts

US

Full

HK

Full

A-shares

Full

Crypto

N/A

Financials

GET /financials/*

US

Full

HK

Full

A-shares

Full

Crypto

N/A

Filings / announcements

GET /filings/sec | /company/announcements

US

Full

HK

Full

A-shares

Full

Crypto

N/A

Ownership / capital flows

GET /ownership/institutional-holdings | /ownership/insider-transactions

US

Full

HK

Full

A-shares

Full

Crypto

Planned

News

GET /news

US

Full

HK

Full

A-shares

Full

Crypto

Planned

Source model

Make every agent answer citeable by design.

FinData4AI is not just another market-data endpoint. It packages financial data as context agents can retrieve, compress, cite, and audit.

Normalized fields

Financials, prices, and disclosures are returned as agent-ready JSON fields.

Source boundaries

Responses preserve source type, publication date, links, and market notes.

Compact context

Retrieve the fields and excerpts needed for the task instead of stuffing full documents into prompts.

Response shape

JSON
{
  "market": "cn",
  "ticker": "600519",
  "dataset": "income_statement",
  "period": "annual",
  "fields": {
    "revenue": { "value": 150560000000, "currency": "CNY" },
    "operating_margin": { "value": 0.689 }
  },
  "sources": [
    {
      "type": "exchange_announcement",
      "title": "Annual report",
      "published_at": "2026-03-28",
      "url": "https://..."
    }
  ],
  "market_notes": [
    "A-share disclosure fields can vary by exchange and report type."
  ]
}

Agent integration

Turn financial research jobs into callable tools.

Start with server-side REST calls, then expose only the validated tools, market coverage, and response shapes your agent should plan with.

Tool schemas match documented request examples
llms.txt points agents to docs and coverage notes
Market filters keep unsupported tools out of plans
Server-side API keys stay outside prompts and browsers

REST

curl "https://api.findata4ai.com/financials/income-statements?market=hk&ticker=00700&period=annual" \
  -H "X-API-Key: fd4a_..."

OpenAPI

const schema = await fetch("https://www.findata4ai.com/openapi.json");
// Generate typed tools for your agent runtime from the same contract.

MCP tool shape

{
  "name": "get_income_statements",
  "inputSchema": {
    "type": "object",
    "properties": { "market": { "enum": ["us", "hk", "cn"] }, "ticker": { "type": "string" } },
    "required": ["market", "ticker"]
  }
}

Grounding pays for itself

Smaller context. Cleaner answers. Lower inference cost.

Instead of pasting entire filings, announcements, and news articles into prompts, agents retrieve only the fields and excerpts they need.

large context

Ungrounded

Whole documents, copied tables, stale snippets, and unclear source boundaries.

targeted context

Grounded with FinData4AI

Market-aware endpoint calls return structured data, source links, and compact context.

Controls

Designed for production workflows.

API key isolation

Create, rotate, and revoke keys from the dashboard while usage stays attributable.

Usage and billing visibility

See request logs, credit balance, rate limits, and subscription state in one control plane.

Source transparency

Expose provenance, market notes, limitations, changelog, and status surfaces as the catalog grows.

Data privacy controls

Manage prompts, API keys, and usage data with clear workspace controls.

Connect your first financial agent in minutes.

Start with sample requests, read the market-filtered docs, then create an API key when you are ready to build.