关于

公开备案是开放的。
我们让他们说一种语言。

这里涉及的每家公司都已经向监管机构免费披露了其数据。缺少的不是数据，而是 62 市场的通用结构、十几种会计标准以及同样多的语言和门户。

OpenFilings 是一个位于瑞士的小型远程优先团队，构建了一次读取文件并使其可随处查询的基础设施 - 文件和 Atom (RSS+)、REST API 和用于 AI 代理的 MCP。

我们的使命

在 OpenFilings 上构建你自己的分析师。

把开放数据交给所有人。让基本面成为全球分析师的共同资产。

用连接 OpenFilings 的智能体补强你的分析，或在同一批披露上自建。MCP 和 API 接到你已经在用的工具。

人人可用的开放数据

直接来自 SEC EDGAR、EDINET、ESEF、CNINFO 等监管机构。每份披露保留来源 URL。

基本面成为公共品

GAAP、IFRS 与亚洲准则下同一套 KPI 名称。营收、利润率、现金流，可跨市场比较。

披露一出即达

监管机构一发布，你的智能体通常可在 60 秒内取到。浏览器、RSS+、REST 与 MCP。

连接你的智能体

MCP 和 API 接到 Claude、Cursor，或你自建的分析师智能体。业绩会、披露与标准化财务，在你现有的栈里。

为什么存在这个

覆盖两个市场的分析师将同一工作重建两次

Regulatory disclosure is a solved problem in each market individually — SEC EDGAR, EDINET, ESEF, and the rest all publish reliably. The gap opens once you look across markets: an analyst or fund covering US tech and Japanese suppliers in the same book ends up reconciling GAAP labels by hand, rebuilding parsers per exchange, and risking a comparison between numbers that were never meant to line up.

We built OpenFilings to do that reconciliation once, in the open, so it doesn't have to be redone inside every fund, every terminal, and every side project that needs it.

Methodology

从监管机构备案到黄金 KPI

1. Source directly from the regulator

Every filing is fetched from the official registry — SEC EDGAR, EDINET, ESEF, CNINFO, and the rest — never from a third-party aggregator or a model's training data. The source URL is kept on every filing record.

2. Extract structured data, not OCR guesses

Where a filing carries XBRL or another structured tag set, we extract from that. Otherwise, parsers built per-market read the actual document structure. Numbers are traceable back to the line they came from.

3. Normalize across accounting standards

Revenue, operating income, net income, and the rest of the Golden KPIs use the same field name whether the underlying standard is US GAAP, IFRS, JP GAAP, or K-GAAP — figures are not restated, only made comparable.

4. Verify, don't approximate

Extracted figures are checked against the filed annual report before being marked reliable. Red flags (dilution, margin compression, cash-vs-earnings divergence) are computed from that verified data, not inferred by a language model.

来源，按市场

62 注册表，一种数据模型

United States

SEC EDGAR

Japan

EDINET (FSA)

Hong Kong

HKEX

Taiwan

MOPS

Israel

MAGNA (ISA)

United Arab Emirates

ADX / DFM

Saudi Arabia

Tadawul

South Korea

DART (FSS)

China

CNINFO

European Union

ESEF / ESMA

United Kingdom

Companies House

Brazil

CVM

Canada

SEDAR+ / EDGAR (cross-listed)

Australia

ASX

India

NSE

Singapore

SGX

每份文件都链接回其源文档——我们从不重述某个数字，而是保留其来源的踪迹。

Who 的幕后黑手

一个小型、远程优先的团队——总部位于瑞士

OpenFilings is built and operated by a small team, not a large data vendor. We run the product ourselves — the parsers, the API, the dashboards — and we're reachable directly.

Questions about a specific figure, a missing filing, or how a KPI was computed? [email protected].

自己探索数据

无需注册即可浏览公司、文件或 Atom (RSS+) 源。

探索公司 →开发者文档 →加入我们 →

公开备案是开放的。我们让他们说一种语言。