# BrowsePDF

> BrowsePDF is a free browser-based PDF toolkit and a programmatic **Document API** for developers and AI agents. The API converts, merges, splits, watermarks, redacts, compresses, OCRs, renders, and extracts text from PDFs, and composes these into pipelines.

## Document API (v1)

Base URL: https://browsepdf.co/api/v1
Auth: send your API key as a Bearer token — `Authorization: Bearer <API_KEY>`
Model: upload a file → receive a document handle → run operations (each returns a new handle) → download.

- [OpenAPI 3.1 spec](https://browsepdf.co/api/v1/openapi.json): complete machine-readable description
- [Developer docs](https://browsepdf.co/developers): quickstart, auth, examples
- [Full reference for LLMs](https://browsepdf.co/llms-full.txt): every endpoint with request/response examples

## Operations

merge, split, pages (reorder/delete/rotate), copy, watermark, page-numbers, fill-form, extract-text, metadata (free), render (page→image), redact, compress, convert (office→PDF), ocr (scanned→searchable PDF).

## Pipelines

POST /api/v1/pipelines composes operations into a DAG — handles flow between steps with no intermediate downloads.

## Conventions

- Async: slow operations (convert, ocr) and large inputs return `{ "job": "..." }` with HTTP 202; poll `GET /api/v1/jobs/{id}`. Force async with header `Prefer: respond-async`.
- Errors: `{ "error": { "code": "...", "message": "..." } }` with a matching HTTP status.
- Rate limits: per-key; responses include `RateLimit-*` headers.
- Metadata is free; other operations cost credits.