How it works
Oxinion pulls filings, market data, and alternative signals so you never wire up a scraper or a data vendor again.
Unstructured documents become typed, schema-defined JSON — every value carries a citation back to its source.
Hit the REST API, or connect over MCP so Claude, ChatGPT, or your own agents read it directly. No UI required.
Endpoints
| Endpoint | Use when | Output | How it works |
|---|---|---|---|
/classify Classify | Files need to be routed by type before processing. | Best-matching category (10-K / 10-Q / 8-K) with per-criterion confidence. | A fast, lightweight step that routes files to the right pipeline before anything else. |
/split Split | One filing bundles multiple sections or documents (10-K + exhibits). | Page ranges for each section, with confidence scores. | Finds section boundaries so each part can be parsed separately. |
/parse Parse | Structured content from a filing is needed for LLM or RAG use. | Structured chunks with typed blocks, page anchors, and confidence scores. | The core reader — turns each section into structured elements that extract builds on. |
/extract Extract | The fields to pull are defined and typed JSON is needed. | Schema-typed JSON with a source citation on every value. | Runs Parse internally and returns only your schema-defined fields. |
Filings
/v1/filings
SEC filings — 10-K, 10-Q, 8-K — aggregated and parsed into typed, source-cited JSON. Extract only the fields you define.
Market Data
/v1/market
Daily price, RSI, trend, score, and BUY/HOLD/SELL signals plus quality/value/momentum/growth fundamentals across the Global Top 100.
Alternative Data
/v1/alt
Web, sentiment, and other non-traditional signals — collected, normalized, and timestamped so agents can model beyond the fundamentals.
Any AI can pull this data
Fetch structured data over REST, or connect the Oxinion MCP server so Claude, ChatGPT, or your own agents read it directly.
GET /v1/filings/AAPL/10-K?fields=revenue,ebitda,net_income
Authorization: Bearer <api-key>
